Skip to content

ON - Batch Uploads Testing & Data Integration - 2025-10-02

Metadata

  • Date: 2025-10-02
  • Company: ON
  • External Participants: Amanda Mitt (Treasury Analyst), Federico Morando (Treasury Intern)
  • Palm Participants: Emma
  • Type: Customer Call
  • Domain Areas: Cash Forecasting, Connectivity, Categorization, Data
  • Recording: https://tldv.io/app/meetings/68de2a554c40390013c1bdfb/

Summary

Context

Customer call to test the batch/file upload feature with ON's tax and talent payment files. Also discussed data integration architecture and how Palm should connect to ON's data sources (BigQuery vs Kyriba).

Key Discussion Points

  • Testing file uploads with multi-sheet Excel files (tax payments, talent payments)
  • Budget code mapping to Palm categories (one-to-one mapping)
  • Column recognition issues - Palm has column limit in mapping UI
  • Number formatting challenges (European format with commas)
  • BigQuery identified as the preferred integration source for AP/AR data
  • Kyriba's daily delete/replace behavior makes it unreliable as data source
  • Palm's approach: only include future dates in forecast, past data saved for variance analysis

Pain Points

  • Kyriba deletes all forecast data daily and replaces - no historical snapshots
  • Kyriba rejects data if bank account not assigned - ongoing data quality issues
  • Google Sheets formulas don't export properly to Excel
  • Column limit in Palm's mapping UI - couldn't see all columns
  • Number formatting (European comma notation) not recognized
  • Training regional teams when forecast data is missing due to Kyriba deletions

Feature Requests & Needs

  • Support for multi-sheet Excel files (select which sheet to ingest)
  • Support for European number formatting (comma as decimal)
  • Ability to edit saved column mappings (currently locked after first upload)
  • Google Sheets integration for tax and talent data uploads
  • BigQuery direct integration for AP/AR data

Jobs & Desired Outcomes

Job 1: Upload forecast data from files without extensive reformatting Desired Outcomes: - Minimize the manual changes required to file formats before upload - Reduce the time spent aligning number formats and column structures - Increase the success rate of first-time file uploads

Job 2: Maintain reliable forecast data despite upstream system limitations Desired Outcomes: - Minimize data loss from Kyriba's daily delete behavior - Reduce the frequency of missing forecast data when training regional teams - Increase confidence in forecast data accuracy

Domain Insights

  • Data architecture: Dynamics 365 → BigQuery → Kyriba (but BigQuery is cleaner)
  • Kyriba behavior: Deletes ALL forecast data daily and replaces - no smart updates
  • Budget codes: Used in Anaplan, mapped to Palm categories
  • File frequency: Tax/talent files uploaded weekly
  • Palm's forecast handling: Only future dates included; past dates saved for variance analysis but not shown in forecast
  • Duplicate detection: Palm detects duplicate rows (same date, same values)

Action Items

  • [ ] Check column limit issue in mapping UI with engineering
  • [ ] Support European number formatting
  • [ ] Enable editing of saved column mappings
  • [ ] Discuss BigQuery integration in Berlin visit

Notable Quotes

"The data goes to BigQuery... Palm could be here instead of Kyriba." - Amanda

"We're training the regions now... and the data is not there because that day because we delete... it happens so many times." - Amanda

"My ambition here is actually to make it easy for you guys... I want the data that lives in files ultimately should be easy for you to upload." - Emma

"Kyriba rejects [data]. So for example, we have rules based on Kyriba bank accounts. If any of the flows don't have a bank account assigned, it will not be integrated." - Amanda

Full Transcript

Date: 02/10/2025, 09:31

00:00 Amanda Mitt: It's nice. Yeah but I'm a bit cold because I think I was walking around and we had this change of weather here so sorry for my voice, it's a bit annoying.

00:02 Emma Sjöström: Yes, my throat is completely.

00:06 Amanda Mitt: Yeah.

00:06 Federico Morando: Or to. So I understand you

00:08 Emma Sjöström: Oh no.

00:09 Amanda Mitt: Everyone is sick. I think Yeah.

00:12 Emma Sjöström: Oh well, I hope it's just like the one and then like the immune system up. Yeah.

00:17 Amanda Mitt: Yeah.

00:19 Federico Morando: Will get used to it. Just a matter of weeks probably but yeah.

00:24 Amanda Mitt: Yeah.

00:26 Emma Sjöström: I think so.

00:28 Federico Morando: Yeah.

00:28 Emma Sjöström: outflows the highlight or notification is that something I'd review? Recognize.

01:00 Amanda Mitt: like, Is it eating that dark right?

01:04 Emma Sjöström: Yes, yes. Should be shared in.

01:18 Federico Morando: Thank you. Okay.

01:21 Amanda Mitt: Yeah.

01:22 Emma Sjöström: The Random Random Feedback tab.

01:25 Amanda Mitt: Yeah, this is a really good feedback.

01:28 Emma Sjöström: All right.

01:28 Amanda Mitt: Yeah.

01:29 Emma Sjöström: Okay. Do you have any idea what?

01:37 Federico Morando: There may be.

01:37 Amanda Mitt: Like me. See. I think it's

01:48 Federico Morando: I don't go.

01:49 Amanda Mitt: Of like accounts a date funding. I think it can be something like that. So like, for example, I don't know anything about 10 million, probably, as if they if it sees an outgo, maybe just list it or yeah.

02:08 Emma Sjöström: Like a large large one.

02:10 Amanda Mitt: every month to met and Google. But as the company's growing, their mother growing and growing and sometimes you kind of don't really keep track. Yeah, I think maybe that's why what he meant.

03:08 Emma Sjöström: Okay. I'll reach out to Rodrigo and chat to him about it a bit more than But in the meantime, we could look at your files that you would like you.

03:23 Federico Morando: and make the liquidity planning tidy, and try to order all the budget codes, but I don't know Amanda, if you want to like,

05:08 Amanda Mitt: Yeah, I know I think it's maybe. Yeah, we have this good help. Because yeah, it's a mapping basically, the thing categories, as palm basically, but we have some codes for them in caliber and that's what we have in the file. So,

05:24 Emma Sjöström: And that would be the budget code field, or

05:26 Amanda Mitt: Yeah. Yeah.

05:28 Emma Sjöström: and then transform the dock. So you have multiple docks with just like wholesales, for example? and I think what we can do is create, for example, in the near term, some form of mapping like you say between This value and the corresponding palm category. Which should and which should allow you.

18:21 Emma Sjöström: There will be a limit to how many rows you can upload in a single file. But I think yours was around 54,000 lines. And I think up to a hundred thousand should be fine for the near term and then we're going to where we're going to get back to you in the workshop.

31:43 Emma Sjöström: When we meet each other in a week and a half with yeah, concrete ideas, also for

06:39 Amanda Mitt: Yes.

06:41 Emma Sjöström: to figure out the categorization bit. but what I'm hearing now is there is Basically a one-to-one mapping between this and Palm category. Correct.

07:07 Amanda Mitt: Yes.

07:08 Emma Sjöström: because the only thing I noticed when looking at the data a little bit is that there are different Flow directions to the amounts.

07:20 Amanda Mitt: And then we just know it by the plus and minus I think as well. so plus wholesale caching means catching

07:28 Emma Sjöström: I know, I think I was just wanted to ask you what it means that is cache in hope.

07:32 Amanda Mitt: Oh okay, it's probably credit note. I think or something like this?

07:33 Emma Sjöström: Negative.

07:37 Federico Morando: Yeah.

07:37 Emma Sjöström: For like the category.

07:53 Amanda Mitt: Okay.

07:54 Emma Sjöström: It's like the cash in cash out, stuff in palm as well, right? So we did the category itself is either an inflow or an outflow. and if I'm gonna do a direct mapping based on this, I'm just, I'm just

08:03 Federico Morando: Yeah.

08:06 Emma Sjöström: interested in understanding When negative flow, how do we? Yeah.

08:12 Amanda Mitt: but, I have no idea how much we see it. Let me check. Like I can do some checks and see how For example, if it's something that in the end will show that we receive one million more or less than I guess. Doesn't matter. But if it's millions of different and

08:29 Federico Morando: it's

08:31 Amanda Mitt: We would care, I think, but I don't think it is that frequent. I I need to check. I can check this.

08:38 Federico Morando: About.

08:39 Emma Sjöström: And there's a few, but they're not. Yeah.

08:43 Federico Morando: like plasma science? How much do we have to split those? So like in the other tracking sheets, the the formatting. I mean for the, for the data, just to also like every teaser for you when we upload the sheets,

09:13 Emma Sjöström: Yeah. So for now it was in the original scope and we can expedite and prioritize we were aware that our customers you know like you have a sheet with or a file with multiple sheets in it.

09:27 Federico Morando: Yeah.

09:28 Emma Sjöström: Being able to select like in the flow, which sheet you want to. Ingest is something. Is that would that be helpful?

09:36 Federico Morando: Uh, yeah, but like, I wanted also to discuss with you, like in the whole when we try to upload the when we tried the test to test the uploads, actually, even

09:46 Emma Sjöström: Yes.

09:51 Federico Morando: with multiple sheets Palm was able to recognize the events and the accounts and

09:58 Emma Sjöström: Yeah.

09:58 Federico Morando: the data. So I don't know if If it's automatically, recognize the first sheet present in the file or if it's able even to link the the movements to the to the proper Evans, if there's a good formatting on our side, but

10:17 Emma Sjöström: I think I think it's so it looks at the columns. First should we try and upload one of your files? Do you want to drive it and just try and upload it? And I think

10:28 Federico Morando: Oh yeah. Yeah, if you I don't know if you wanted to ask more about the this file or like then we can see

10:35 Emma Sjöström: it's really how we should think about these cases, if and I think If this is truly correctly mapped from your side as caching wholesales.

10:48 Amanda Mitt: many examples. but,

10:55 Emma Sjöström: And and just what I heard from you was like you wanted to check like the total value of those flows, or those amounts and

11:00 Amanda Mitt: Yeah.

11:02 Emma Sjöström: If it's not too much for now, like the near term, we what I'm hearing is we could potentially ignore them.

11:08 Amanda Mitt: Yeah, exactly.

11:08 Emma Sjöström: Yeah. Yeah. So but please let us know what you find out and we'll make sure to find a nice way to solve it, okay?

11:17 Federico Morando: Okay. Okay. Thank you.

11:18 Amanda Mitt: Okay.

11:20 Emma Sjöström: Of course.

11:25 Federico Morando: Wait, let me share. So here I provided a small extract, I don't know if you want to try with both the extract and the original file to see how much can recognize right now. But

11:40 Emma Sjöström: Yeah.

11:44 Federico Morando: here we have the three most important sheets that can be useful because here we have the data In Eden columns that what? That's what we came out with with the tax team. Years are the sources, and here, our accounts and our entities based on these master data. Like bank accounts.

36:12 Federico Morando: And so from here, it should from the first sheet. It should

12:12 Emma Sjöström: All right.

12:18 Federico Morando: only recognize the The the only one accounts, the one that we used to to pay, I mean, and With the Eden columns should be. It should work like as we as well, some issues, but then like by formatting a bit, will I remember that last time that we tested? It was a bit better even then better than they call that we had with Christian discussing the Issues.

12:48 Federico Morando: So wait, let me

12:48 Emma Sjöström: All right.

12:54 Federico Morando: Share my full. Screen.

12:58 Emma Sjöström: Okay. Sorry, I do apologize. We need about 10 more minutes for before we can try the actual

13:04 Amanda Mitt: That's okay.

13:05 Emma Sjöström: I apologize.

13:09 Federico Morando: Okay.

13:10 Amanda Mitt: No, may just have a quick question, pretty good for the tax payment. For example, it's When you tested. Because I think we might have some screenshots as well. I don't know if it worked it, but did you? It recognized the direct text, but the indirect, it didn't right that it's the

13:28 Federico Morando: Oh no, that's yeah, we planned the we have a cold next week because our breed,

13:28 Amanda Mitt: same negotiate. Yeah.

13:36 Federico Morando: tax and then upload it separately and but I wanted to also one understand better with him so that we can a bit think about The design of the process that also like it will, I hope it will. Like also after the when you will come here in Berlin we about some stuff.

31:21 Federico Morando: We can also like I wanted to brainstorm a bit to B coordinated with the other teams, and also have a streamlined process to do that. So that it's it goes smooth. And I want to make it

14:30 Amanda Mitt: I think this one we probably, can you share just the Google sheets of the text one so when I can see

14:35 Federico Morando: Yeah. this is the extract, but this is the actually,

14:41 Amanda Mitt: Yeah, it won. That is the direct text and then indirect tax. Yeah, so I think in our tests it only recognizes direct text, right?

14:49 Federico Morando: Yeah. Yeah, I think so we can try again to verify but

14:55 Emma Sjöström: Do they have different columns or anything? No. Or is it just the account

14:56 Federico Morando: Yeah.

14:59 Emma Sjöström: numbers that were

15:01 Federico Morando: No, no. The columns are the same is just the Yeah, like it's

15:08 Amanda Mitt: Because we didn't split anything. We just like we downloaded the full sheet and we have uploaded it like this. Easiest simplest way possible, just to check, what would happen? And it worked for the first tab for some reason, but It only read the first step which I don't think it's bad.

15:27 Amanda Mitt: It's just what happened.

15:29 Emma Sjöström: Interesting.

15:31 Federico Morando: but let's try with both like,

15:33 Emma Sjöström: Yeah, I'm sorry. We're still deploying the fix to production, so I'll get a ping when it's

15:40 Federico Morando: It's okay.

15:43 Amanda Mitt: Yeah, I wanted to ask Qmr if it's like clear, how Kitty, but is consuming,

15:44 Emma Sjöström: Apology.

15:49 Amanda Mitt: Apar. I don't know if it's super clear.

15:57 Emma Sjöström: Do you mean, Oh, it's consuming from Cariba. Point of Federico. Please wait a

15:57 Amanda Mitt: How?

16:02 Emma Sjöström: little bit. We're still waiting.

16:04 Amanda Mitt: Yeah.

16:04 Federico Morando: Yeah. Sorry. Sorry. I was just

16:06 Emma Sjöström: No. It's like excited.

16:11 Federico Morando: Sorry.

16:12 Emma Sjöström: Don't worry. Um you mean in terms of how caribou fetches the data,

16:18 Amanda Mitt: Yes, because we have this daily process.

16:21 Emma Sjöström: Mm-hmm.

16:22 Amanda Mitt: kill you, but we do some changes. Yeah.

17:56 Emma Sjöström: How do you do? You do the in Korea?

18:00 Amanda Mitt: Yes, we do the rules and caliber.

18:02 Emma Sjöström: All right. And how what would that what does that look like? It's basically like if expected payment date is today and it didn't happen, then you just ignore or because you don't like

18:15 Amanda Mitt: yeah, I can say for example like the hex data here, so this is, Yeah. The data photo I think from 2020 here, right? So That's still open. Yeah, sorry, it's super small.

18:29 Emma Sjöström: I, Back in the days.

18:34 Amanda Mitt: Breaking it. Yeah, so like we exclude all of this but it's still in the system. So we have this kind of data job interview. I can show you how it looks like.

18:44 Emma Sjöström: Gotcha.

18:58 Amanda Mitt: So, one second, let me just open the mapping rules.

19:03 Emma Sjöström: Yeah.

19:07 Amanda Mitt: Okay management. Yeah, I can tell you here, so Can you see? This is the list of all the bench modifications.

19:43 Emma Sjöström: Oh yes.

19:47 Amanda Mitt: Delete forecast, integrated today and of the day. So every end of day, Delete, and it replaces.

20:00 Emma Sjöström: Mmm.

20:01 Amanda Mitt: Um, delete best forecasts for example, let me see.

20:05 Emma Sjöström: And when you say replace delete and replaces, does it move around any dates or is it just that it deletes the ones you're expecting? That didn't happen or like and then you just

20:16 Amanda Mitt: No, it just delete all passwords, it doesn't check anything. It's like not a

20:17 Emma Sjöström: Know.

20:22 Amanda Mitt: smart. It's just delete class data.

20:23 Emma Sjöström: Yep. Can I ask you just like a more high level sanity?

20:28 Amanda Mitt: Yeah.

20:29 Emma Sjöström: Check question.

20:31 Amanda Mitt: Yes.

20:34 Emma Sjöström: Now to either match uploads or automatically just in just all of your open apna are as of today. And with with payment states only in the future, right? Or like expected dates only in the future. Cool. So how we would think about it is that now this is a part of your forecast.

37:44 Emma Sjöström: And whatever happens or doesn't happen. We actually, we still save the data but they become like a past forecast so they're no longer they're automatically. No they're no longer part of what we're still expecting because they expected payment date passes.

21:33 Amanda Mitt: Yeah.

21:34 Emma Sjöström: And why we? Kind of like to save it on our end is because then invariance analysis, it's easier to show what was the reason of the variance. Okay, we expected this customer to pay their invoice. They didn't

21:47 Amanda Mitt: Yeah. That's really yeah, it's ideal.

21:49 Emma Sjöström: Is that a fair assumption? It's we're still laying out the foundation, because we do have ambitions around variance analysis to service a lot easier. Like, all the wise behind you had this like Gap for example. And in the future, of course, it would be very cool to even explain, You know.

38:34 Emma Sjöström: Oh, the implode came later and maybe it's because of this one that didn't come as expected. So we would still like to save all data, but we would only in the forecast include anything that's like still has a future date.

22:23 Amanda Mitt: Yeah.

22:24 Emma Sjöström: Is that?

22:25 Amanda Mitt: That's good. No, yeah, I think that's the idea way. Like, for example, in this short term focusing, anyway, you don't even have various analysis And you would have to do some processing Excel, for example, download the data compare, day by day, you know?

22:39 Emma Sjöström: Yeah, I gotcha.

22:41 Amanda Mitt: automate

22:54 Emma Sjöström: I understand.

22:55 Amanda Mitt: Not really. So for example, if I see There's a very analysis worksheet, but we never even. I don't even know how it like, It would it would probably delete. It's for sure there's not storage the data.

23:07 Emma Sjöström: oh,

23:12 Amanda Mitt: It's probably from the data from today, you know?

23:16 Emma Sjöström: Mmm.

23:18 Amanda Mitt: yeah, because the only thing they do incredible is like they they counted as estimated forecast versus actuals but that's not If the data changes daily, then that's not really. A very analysis It's just, you know, we know what I mean. Like, in Inform, You have some sort of snapshot, I think

23:38 Emma Sjöström: Yeah, and we would do currently we do. Yeah, we used the last known for a cost for the like, past week for. So the freshest for us is always used in the

23:45 Amanda Mitt: Yeah.

23:49 Emma Sjöström: week, we create a new forecast in terms of the machine learning. But in terms of the actuals or the committed cash flow that you would upload, that wouldn't change, we wouldn't change that but we would not include stale data in the

24:04 Amanda Mitt: Okay.

24:08 Emma Sjöström: forecast. Like Okay, you expected it to happen today. It didn't then it's part of the history and it's part of the variance analysis, essentially.

24:16 Amanda Mitt: Yeah, that's good.

24:18 Emma Sjöström: Cool. Because I think it would be super interesting to understand. Also in your bigquery, it comes originally from Dynamics, the Apar

24:28 Amanda Mitt: Yeah. Yeah.

24:29 Emma Sjöström: You also ingested into Bigquery?

24:32 Amanda Mitt: Yes.

24:33 Emma Sjöström: And then on top of Bigquery, you have Hex. Is that correct? Or is it also?

24:38 Amanda Mitt: No, sorry. I remember I had this to do of lifting all of the Sources. No, it's Let me think I have this video once. But it's Hex is just for us to see the data, it's not part of nothing like on

24:55 Emma Sjöström: Yeah.

24:55 Amanda Mitt: this setup. It's just dynamics bigquery caliber so

25:00 Emma Sjöström: Yeah.

25:01 Amanda Mitt: Yeah, this center is Bigquery. I think where they

25:05 Emma Sjöström: and then,

25:06 Amanda Mitt: Treat the tables. A bit. Yeah.

25:07 Emma Sjöström: and then bigquery you have already the budget code as well, or is that include like

25:13 Amanda Mitt: Yeah. That in Bigquery.

25:16 Emma Sjöström: Yeah, basically, all fields we need are in bigquery. We have the expected

25:16 Amanda Mitt: Yeah.

25:22 Emma Sjöström: payment date, we have the code, maybe the cash in cash out a code as well, the

25:23 Amanda Mitt: Yes.

25:28 Emma Sjöström: amounts,

25:29 Amanda Mitt: Yeah.

25:30 Emma Sjöström: Descriptions, like all of those that's in Bigquery already.

25:34 Amanda Mitt: Yeah, nothing from Kidiba.

25:36 Emma Sjöström: Yeah.

25:36 Amanda Mitt: so the data team did, I think how you call it is not matter data maybe but yes I'm part of matter data I think with the budget called and yeah

25:47 Emma Sjöström: so,

25:47 Amanda Mitt: for for example, some rules as well. So you yeah, nothing from

25:56 Emma Sjöström: So then, I think, actually the most stable, just we'll talk more in Berlin, but I think the most stable form of integration would be with Bigquery.

26:05 Amanda Mitt: Yes.

26:06 Emma Sjöström: Would be the cleanest and mostly in terms of the data.

26:10 Amanda Mitt: Yes.

26:11 Emma Sjöström: Actually.

26:14 Amanda Mitt: I find this. Dashboard.

26:23 Emma Sjöström: Thank you for talking about this. This also super helpful because it helps us understand how we can get that. We want to automate the AP and AR for you. It's just a matter of how to do it. But what I'm hearing now is Bigquery is the possible option.

26:40 Amanda Mitt: Yes.

26:41 Emma Sjöström: from your, from your side, or outside, how to deal with that. Yeah.

26:57 Amanda Mitt: Yeah. Yes, writing this is why well like why can't we just upload as a Google Sheet? Because we do all of these types of swelling cable. So I think that's

27:03 Emma Sjöström: Yeah.

27:06 Amanda Mitt: what we're a bit concerned about

27:08 Emma Sjöström: Google Sheet integration is also high on your wish list or

27:13 Amanda Mitt: I think for tax and talent. Yes, but not for Apar. Yeah.

27:17 Emma Sjöström: No.

27:19 Federico Morando: Not.

27:20 Amanda Mitt: so,

27:21 Emma Sjöström: It would work in the near term to export Google and upload in the near term.

27:28 Federico Morando: So, could you?

27:29 Amanda Mitt: Sorry it cut a bit. Yeah.

27:31 Emma Sjöström: For the file uploads feature to get the tax and talent data into the forecast of Palm. Would it work in the more near term to still export a file from Google sheets and upload it Is better for now.

27:43 Amanda Mitt: Yeah, we don't we don't see many changes for now in that. Yeah.

27:44 Federico Morando: so, for

27:46 Emma Sjöström: Yeah. Okay. Okay, what is this? Oh yeah, I've seen this before. I think.

27:48 Federico Morando: Yeah.

27:53 Amanda Mitt: Okay, so yeah, here's like Nature 65 and bigquery.

27:54 Emma Sjöström: But he's Hmm.

27:59 Amanda Mitt: And then to kiriba.

27:59 Emma Sjöström: Mmm.

28:01 Amanda Mitt: so, it doesn't matter but basically like The data goes to just be query, NOx. Sorry.

28:12 Emma Sjöström: Don't worry.

28:13 Amanda Mitt: Sorry I'm really like I don't know what's going on. I have it's been one weekend

28:14 Emma Sjöström: Yeah.

28:17 Amanda Mitt: like this and just getting worse. These are 65, Anna Plan. Everything goes to Bigquery.

28:25 Emma Sjöström: Yeah.

28:26 Amanda Mitt: Yeah.

28:28 Emma Sjöström: That's good news.

28:30 Amanda Mitt: And then Palm could be here like instead of cable. Yeah.

28:34 Emma Sjöström: Yeah, if we could get as much as possible directly from Bigquery. Now, we

28:38 Amanda Mitt: Yeah.

28:40 Emma Sjöström: Much more stable.

28:42 Federico Morando: Cuz it's for you because it's a cleaner data there, right?

28:43 Emma Sjöström: You know. Yes, and

28:47 Federico Morando: Okay.

28:50 Emma Sjöström: Yeah.

28:55 Federico Morando: Nice. This is Yeah, I think like, basically for talent and tax both direct and indirect that would be the initial star where we can also like try to better shape the Things also with Palm for also they maybe the APA are more immortal for the future. I mean, that would be A good thing to do to also understand this.

29:23 Federico Morando: It's I mean this is super old lies, you were saying This is super helpful because we can also align a bit towards Your visit on some stuff so, yeah.

29:38 Emma Sjöström: 100%. Hey, good news to fix this out. So we can try the file upload flow. Now if you want to federico, unless there was anything more you wanted to say here, Amanda?

29:51 Amanda Mitt: Manualable authority VR and yeah clarified.

29:58 Emma Sjöström: No. I get like a post increase. You're seeing that.

30:01 Amanda Mitt: So yeah.

30:05 Emma Sjöström: All the rules that you have to maintain, but yeah.

30:07 Amanda Mitt: yes, yes, it was a lot of work and I can also share some more files where we like,

30:12 Emma Sjöström: Yes.

30:12 Amanda Mitt: It was a lot of work just creating this table, you know.

30:16 Emma Sjöström: I bet.

30:18 Amanda Mitt: Yeah.

30:20 Emma Sjöström: I bet let's try and make sure that you don't have to do a lot of more work to get it into.

30:24 Amanda Mitt: And it's like the video is always talking to the data team because it fails constantly and caribou rejects. So for example, We have rules based on cleaver bank accounts. If any of the flows don't have a bank account assigned, it will not be integrated. So it's a very like Now it's getting more robust but it's still very ongoing.

30:46 Amanda Mitt: It's not ideal, it's not like we're super confident with the process. Um, still Yeah. So just like you

30:51 Emma Sjöström: Yeah.

30:53 Amanda Mitt: Yeah.

30:55 Emma Sjöström: I understand, I still I will I will be transparent and say that I do expect we would need some that we would need some notion of understanding which account to ingest to on our side as well.

31:06 Amanda Mitt: Okay. Yeah.

31:07 Emma Sjöström: And maybe there's something clever we can do about the things that fail. Yeah. Yeah.

31:16 Amanda Mitt: Yeah.

31:17 Federico Morando: Yeah.

31:18 Amanda Mitt: Okay. That's yeah good. I think it was nice here. Discuss it. Thanks.

31:25 Emma Sjöström: but feel free to share if you find any like example of such data like, Oh, this

31:26 Amanda Mitt: All right.

31:30 Emma Sjöström: is A batch of data and it failed because these two didn't have bank accounts associate like, just share it with us and we will have a thing.

31:39 Amanda Mitt: Yeah, and because it changes daily. It happens so many times like We're training the regions now, for example, we go and we have a session with on China accounting team because they also the Treasury work, they need to check the forecast and then the data is not there because that day because we delete and it's that they, you know, it happens so many times.

31:59 Amanda Mitt: So, Yeah, just letting you know.

32:02 Emma Sjöström: I expect that wouldn't be as big of an issue if it was from Bigquery directly and we just look for duplicates instead from Bigquery and ignore those and then we ingest the rest.

32:11 Amanda Mitt: Yeah.

32:11 Emma Sjöström: Yeah. Okay, cool. right, let's

32:15 Amanda Mitt: Or like like if the data didn't come and maintain the maintain something, you

32:16 Emma Sjöström: Yeah.

32:19 Amanda Mitt: know like don't don't believe everything like but it we cannot do it so yeah. Okay.

32:24 Emma Sjöström: Yeah, no, I did. I have heard about how Kariba does. In fact, do that? Like

32:25 Amanda Mitt: Sorry.

32:30 Emma Sjöström: daily deletion a little bit by behavior, which also makes it hard for us to in

32:31 Amanda Mitt: Yeah.

32:35 Emma Sjöström: turn rely on Kariba, because we understand just from that behavior that it's going to be tricky to Yeah, into a correctness of the data. Yeah.

32:48 Amanda Mitt: Yeah, it's not ideal.

32:49 Emma Sjöström: Not ideal. but I think there's big hopes, I think and yeah, if we can get access to Bigquery, We'll figure it.

32:59 Amanda Mitt: Yeah, thanks. Sorry for the video.

33:02 Emma Sjöström: Oh, it's so super useful. Thank you.

33:02 Federico Morando: Know. Sorry. Yeah. So I also to understand better from both sides so it's no word and yeah so

33:11 Emma Sjöström: Yes.

33:14 Federico Morando: should we Try to upload.

33:18 Emma Sjöström: Yes, it should be fixed now.

33:20 Federico Morando: Okay.

33:21 Emma Sjöström: We had a small bag, I just didn't want to us to run into that. The first thing we did,

33:27 Federico Morando: Okay. so,

33:32 Emma Sjöström: Maybe we know the pay or I think it's fine actually, but just just to be sure.

33:36 Amanda Mitt: Oh yeah.

33:39 Federico Morando: Sorry.

33:40 Emma Sjöström: as maybe just reload the page and and just

33:43 Federico Morando: Oh yeah. Yes. So that you

33:44 Emma Sjöström: I don't, it should. Shouldn't be an issue, but you never know.

33:49 Federico Morando: Yeah, so we'll try with this. Be cash out tax.

34:00 Emma Sjöström: is this now the file with all the sheets in it or just one sheet

34:05 Federico Morando: This one is the one with all the sheets.

34:08 Emma Sjöström: Gotcha. Yeah.

34:09 Federico Morando: So that we can see what's like, what was wrong?

34:13 Emma Sjöström: Sounds good.

34:16 Federico Morando: All right, this second that I have that. Cool thing. Okay.

34:22 Amanda Mitt: If you want, you can close it and it won't.

34:25 Emma Sjöström: Just a

34:28 Federico Morando: And so, okay, backup account. Even account is right. So Lamont, payable amount was right description and note, 2, and you dated validate. So it recognized, the The. Categories.

34:44 Emma Sjöström: Mm-hmm.

34:44 Amanda Mitt: because,

34:46 Federico Morando: Confirm but it says invalid. Now we see why. Okay. Bank account, even be able.

35:00 Emma Sjöström: So it's something about the bank account. I wonder these events are those actually your accounts or the tax authority account?

35:11 Federico Morando: So like there are, we are these issues so that basically for the direct tax. most of the payments were directed a bit, so they weren't like, registered to aiban and also like, for example, From the for the talent tops. We for the talent tops file. We didn't have actually our accounts and that that's what we try, we were trying to solve.

35:45 Federico Morando: And so I think these let me check on in the other file. So, Should be. Yeah, probably they

35:55 Amanda Mitt: Nobody, I think we need to test after we made the changes that we proposed not with the original one.

36:02 Federico Morando: No, no, but it's with the changes.

36:04 Amanda Mitt: Okay.

36:06 Federico Morando: Yeah, it's just the hidden column at the end.

36:10 Amanda Mitt: Okay.

36:12 Federico Morando: You see here?

36:14 Emma Sjöström: Oh yeah.

36:14 Amanda Mitt: I didn't. Yeah. Then I think it's probably just changing maybe supplier account

36:15 Emma Sjöström: that's,

36:20 Amanda Mitt: or something like this.

36:22 Federico Morando: yeah, I don't know that like, for example here, If like I upload the sheet and I want to change the categories because it it's just cause automatically, right?

36:35 Emma Sjöström: Yeah.

36:36 Federico Morando: But if, for example, I change bank account, even to, for example, this column, which was in bank account, even Treasury only. It should be able to recognize it

36:45 Emma Sjöström: Hmm.

36:47 Federico Morando: right?

36:50 Emma Sjöström: And so this is something that we are going to fix because we need to unlock these fields so that you can change the mapping here.

36:56 Federico Morando: Okay. Okay. Because I remember like, for example, when I was trying to upload the talent, tops one, I was able to To choose the columns through the To the, through the selection here. So, Or you can do it once the palm just not is just not able to recognize it.

37:19 Federico Morando: So you have to allocate it properly manually or

37:24 Emma Sjöström: So how how it works is like the first time you upload a file. We recognize the file is by columns essentially the names of the columns.

37:35 Federico Morando: Yeah.

37:35 Emma Sjöström: And the first time you do that, we have sort of and we have an AI that guesses and initial mapping. So based on the column contents and the palm fields, you see they're on the right. It kind of guesses for you, what the mapping should look like. But then you can always Amanda.

37:53 Emma Sjöström: You can change that if it gets. If due date is wrong, for example, it should be payment date or some other column, you can Adjust it. And then what happens is after that, we kind of, we save these mappings for that file.

38:09 Federico Morando: Okay.

38:10 Emma Sjöström: But unfortunately, right now they're locked in, we are going to release so that you can change this.

38:17 Federico Morando: Okay.

38:18 Emma Sjöström: Because mistakes happen. And this is the thing that we're well aware that we need to like release so that you can actually go back and change the mappings. Did that make sense?

38:29 Federico Morando: Okay, yeah, yeah, absolutely. And So, maybe we can try because of that to To try with the extract. And also, I don't know because if now it recognize bank account, even for account,

38:50 Emma Sjöström: Mm-hmm.

38:50 Federico Morando: Maybe bus by just renaming the columns.

38:54 Emma Sjöström: Try it. Yes. I that's how I expected should work.

38:57 Federico Morando: Yeah, like not. Not maybe. With the original file, because what we came up with for example with the charger from the tax team was to okay select and put the accounts and all the data source at the end. So that who like they use this file since a while.

39:18 Federico Morando: So we didn't really want to mess up there for Martin if they're used to Use it in a comfortable way and it's comfortable for them. So yeah. But we can also like maybe try with an extract by filtering it to do it. I don't know if it's If it becomes.

39:39 Federico Morando: too much manual work to do on the long term, but I mean,

39:46 Amanda Mitt: I think it's fine, but I can also try with Georgia. Maybe just add supplier or pay like Directed between account, something like this. But I think if we have to change it you like probably we will upload this once a

39:59 Emma Sjöström: One.

40:02 Amanda Mitt: week or even less. So I think it's okay if we need to change it.

40:05 Federico Morando: Okay. Perfect. So that would, that could be an option. And yeah, but we can try also with this extract, which is only the direct tax payments, and we have the events in the first sheet, so we can see if it's if it recognize it or not.

40:26 Emma Sjöström: Let's do it.

40:32 Federico Morando: Okay, let's Start again. There. Okay. Just

40:43 Amanda Mitt: I was like,

40:45 Emma Sjöström: Perfect being on those.

40:48 Federico Morando: Always.

40:48 Amanda Mitt: It's, it's so weird. Every started mine yesterday. And now it's asking again,

40:49 Emma Sjöström: Yeah.

40:54 Amanda Mitt: it's you.

40:55 Federico Morando: Yes, yes, the same for me to, but

40:59 Emma Sjöström: Annoying.

41:01 Federico Morando: Yeah, the know why? But okay. Okay.

41:14 Emma Sjöström: Here you go. Here, you have an option now to set the correct.

41:17 Federico Morando: and now I can We, but what she Mmm. But this is the date, I think.

41:33 Emma Sjöström: Yeah.

41:34 Federico Morando: Because wait.

41:36 Amanda Mitt: These are the columns now.

41:36 Emma Sjöström: there was Sherry only is the Treasury only column not in there.

41:40 Federico Morando: Yeah, that's what I was wondering because it goes. I don't know if wait. Because it should be after Jnl.

41:51 Emma Sjöström: Yeah. On.

41:53 Federico Morando: if we oh if we open the source file that you like the export the file export

41:58 Emma Sjöström: Maybe.

42:04 Federico Morando: directly from your downloading download folder, To see that it's really there.

42:17 Emma Sjöström: like, Yeah, let me check you. Usually use Google sheets.

42:18 Amanda Mitt: Oh, it's

42:21 Federico Morando: And this happens. Sometimes it's so annoying with Google sheets.

42:22 Emma Sjöström: Because it doesn't match the formula on Excel.

42:25 Amanda Mitt: A.

42:26 Federico Morando: Okay.

42:27 Amanda Mitt: Okay.

42:32 Federico Morando: Let me think.

42:34 Emma Sjöström: Hmm.

42:37 Federico Morando: Hmm.

42:39 Emma Sjöström: Can you see it, right? I'm sharing the full screen, okay?

42:42 Federico Morando: Yeah.

42:48 Amanda Mitt: Now, it's normal. This happens a lot with Google Sheet and Excel What if you download it here? like,

42:56 Federico Morando: And if I sorry, if I write the formers here, like

43:00 Emma Sjöström: Maybe.

43:00 Amanda Mitt: Yeah, you can just

43:02 Federico Morando: Let me try.

43:03 Amanda Mitt: But it this will no of course. Yeah this is what I usually do, but

43:08 Emma Sjöström: It was me export. Try it out.

43:08 Amanda Mitt: It was to be a bit annoying. Um, but if you don't know that CSV, do you think it like wait.

43:13 Federico Morando: I, it's Yep, now. Sorry.

43:21 Emma Sjöström: Only. it's

43:25 Amanda Mitt: Which one elements are.

43:27 Emma Sjöström: It's very useful for me to just see, even this part of the process, right? It's it's a little bit cumbersome to say

43:33 Federico Morando: Yeah. Yeah.

43:35 Amanda Mitt: Yeah, it's great. Like you don't even expect this small thing.

43:39 Emma Sjöström: No.

43:41 Amanda Mitt: You've also said CSV, or you said another format?

43:45 Federico Morando: No, no Csv.

43:47 Amanda Mitt: No, no Emma. You said it Csp.

43:49 Emma Sjöström: It should work.

43:50 Amanda Mitt: Okay.

43:52 Emma Sjöström: Um, so CSV and Excel. How about now did it include?

43:58 Amanda Mitt: No.

43:58 Federico Morando: M, No.

44:00 Emma Sjöström: What?

44:00 Amanda Mitt: Let me go away. Wait, one second. Okay.

44:06 Emma Sjöström: That is strange. It should really. I mean, even if the values are not Okay, within the column, the column header is there.

44:16 Federico Morando: Yeah, but

44:21 Emma Sjöström: Strange.

44:22 Federico Morando: Yes. Mmm, because wait.

44:31 Emma Sjöström: What if sorry would you mind trying again? This is going to sound super weird Federico but close.

44:36 Federico Morando: No, no, no. We're here for this. So no, no worries.

44:41 Emma Sjöström: Close this week.

44:43 Federico Morando: so,

44:44 Emma Sjöström: And and start again, upload the same file.

44:46 Federico Morando: yeah, let me Okay.

44:55 Emma Sjöström: And add four cost after a spreadsheet. And now we're not going to India, speak another category, so not tax category, but some other outflow.

45:06 Federico Morando: Okay.

45:08 Emma Sjöström: Pay, we're not going to ingested for now. We're just gonna test if the column shows up because it's a little bit strange. Upload. I wonder if there's some setting. Man. Okay. it's

45:32 Federico Morando: But I see.

45:33 Emma Sjöström: Yeah, there's no not something else behind after J&l I think. But

45:37 Federico Morando: There is yeah. It's interest among Julie category. well, let me Can he as a limit on the columns that it can recognize?

46:03 Emma Sjöström: I will need to check with the engineers but there shouldn't be but that's very yeah, could be.

46:10 Amanda Mitt: You know, a copy paste. What if you copy paste the formula? I think it's

46:16 Emma Sjöström: Move it.

46:22 Amanda Mitt: because I was looking and I think it's like when you convert Google sheets to excel, if it's not a if it's a Google formula Then that's what, why Excel?

46:29 Federico Morando: Yes. Yeah. Also like Yeah we can also do the formula again but I wanted to see if like,

46:41 Emma Sjöström: Yeah, I know it makes sense. Let's try and like

46:43 Federico Morando: was just a simple thing of like, You know, it's moved. and we got it on, column you Alright, let's let's try again. Okay.

47:02 Emma Sjöström: Nice.

47:12 Federico Morando: We were saying payable so that we

47:14 Emma Sjöström: Yeah, let's try it out so because we do sort of save stuff related to the category as well. but,

47:28 Federico Morando: Yes.

47:29 Emma Sjöström: Yes, okay. So there seemed you're very right. Good good.

47:33 Federico Morando: We got it.

47:35 Emma Sjöström: Sometimes you in fact have a limit on how many columns we can use, which is strange, but I'll bring it up with the engineering team and see what's going on there.

47:43 Federico Morando: or maybe just, it was a just about, like, like out shows because I was trying to go down, but it maybe it was at the end of the

47:55 Emma Sjöström: Oh well, did it feel like hard to was it not easy to? Okay. Okay.

47:59 Federico Morando: Yeah, the

48:02 Emma Sjöström: Good, thank you. We'll look into that.

48:05 Federico Morando: Okay, and then so with like yes, the same thing. What's

48:13 Emma Sjöström: oh, it's tricky to

48:16 Federico Morando: Yeah, but

48:17 Emma Sjöström: Yeah.

48:19 Federico Morando: Wait. That would be total amount payable. The scription and payment date to be filled by a P. Just let me check if it's correct. It's due date. I think. Yeah. Let's do date. Alright.

48:43 Emma Sjöström: What? Sorry, just a question. What's the difference here between the due date and the AP payment date? Or like, what is it?

48:49 Amanda Mitt: I think it's when the AP team actually creates the payment, like, they just confirmed it. No. Or

48:56 Emma Sjöström: But it should be. Forward look. Okay. So, so what, which one do you want in your forecast essentially?

49:04 Amanda Mitt: Yeah, the fingers. Can we see? I don't let me see some examples Distribution. Do they have the same understanding for the record? You think it's different meaning

49:14 Federico Morando: So, that's

49:15 Amanda Mitt: They think it means something different.

49:20 Federico Morando: And we can check it, but wait. Let me go on the From the original file.

49:30 Amanda Mitt: so I see, for example,

49:33 Federico Morando: So, due date was year. okay, for example, let's go to 22 September like this.

49:50 Amanda Mitt: Yeah, no, it's when they pay it. They feel it up to confirm.

49:53 Federico Morando: Yeah.

49:54 Emma Sjöström: okay, so it's like after the fact sort of

49:56 Amanda Mitt: yeah, then

49:57 Emma Sjöström: Got. Interesting.

50:00 Federico Morando: This one.

50:01 Emma Sjöström: Okay.

50:03 Federico Morando: I don't know if it's better to actually have when the payment is made, but the thing is that probably for example, for yeah, we should wait every time for maybe the payment day to be filled and maybe if it's not done like Already. Yeah, I don't know.

50:30 Amanda Mitt: Now we can ask Kata, I think, what exactly this means.

50:34 Emma Sjöström: Yeah.

50:36 Federico Morando: Yeah, but

50:38 Emma Sjöström: Let's go ahead and see if it actually maps.

50:42 Federico Morando: Yeah.

50:43 Amanda Mitt: Yes.

50:43 Emma Sjöström: I think it might be. A value dates in the past now but let's have a look at the

50:51 Federico Morando: Yeah.

50:52 Emma Sjöström: K.

50:53 Federico Morando: Well, which is now is here is not good.

50:56 Emma Sjöström: That's it. Find the account at least.

51:02 Federico Morando: Why?

51:03 Emma Sjöström: I think it's having you with a number formatting.

51:07 Federico Morando: Ah. Okay. You mean the comma or the no?

51:12 Emma Sjöström: I would suspect it's that. Yeah, the up up comma there.

51:16 Federico Morando: Yeah.

51:17 Amanda Mitt: Also, it's positive. No, it's a cash out or it's a machine. Yeah. But things with negative. Yeah.

51:23 Emma Sjöström: Yes. And there is another one that has a different direction.

51:31 Federico Morando: I'm sorry, I was trying to check if Here. But why, in case some it's true. Some are recognized and some are not for

51:34 Emma Sjöström: so,

51:38 Federico Morando: example.

51:40 Emma Sjöström: But the one there with the amount is valid. Yeah, that's strange. They're the same. Does it have something else on it? That's invalid. Like that line.

51:51 Federico Morando: Hmm. Sorry, trying to Say something. Everybody account or not but it recognized the events that's good.

52:03 Emma Sjöström: Yeah, exactly. So I think now it's just the Number formatting, but let me, can we just quickly look at the original file. Again was the numbers. Formatted.

52:15 Federico Morando: Yes.

52:16 Emma Sjöström: This. Yeah, it's okay. And this is this fairly standard to how you guys want to format numbers.

52:26 Federico Morando: No. Like we can discuss it with the teams because also like I remember talking

52:26 Amanda Mitt: No.

52:32 Federico Morando: about this with Christian and he said that like actually Palm could have recognized it automatically but it's not an issue at all. Like we just have to To standard the format, which is the best to be recognized with with the teams. So,

52:54 Emma Sjöström: Yeah, all right.

52:55 Federico Morando: Yeah, they would. We can do that.

52:56 Emma Sjöström: Yeah. If you want to do that and you can let us know. I will make sure to support like the number formats that you use.

53:07 Amanda Mitt: We don't need it to be negative. No. I think we will need to create a column and make it negative. I think that's

53:11 Federico Morando: That's right.

53:16 Amanda Mitt: what Christian said that we cannot do anything about, like they cannot do anything about right.

53:20 Emma Sjöström: Ah, and not, so sure about that. Let's

53:24 Amanda Mitt: Okay.

53:25 Emma Sjöström: If we can start with, just at least having the numbers formatted in a way that makes sense to you. And if this is the way that makes sense, You know, let us know. Well. the key, the main issue when We have them. Because the poem categories have the restriction on the flow direction.

00:00 :

53:48 Federico Morando: Okay.

53:49 Emma Sjöström: It's when we start getting a file that has mixed that, like some numbers are negative, some numbers are positive. Then it becomes a bit more complex on our side because then we have to go through each number of the file. And I think that's fine for smaller files. To be honest, engineers, might have different opinion, but I it's it's more an issue for the larger data set like the AP AR.

54:19 Emma Sjöström: It will take more time to ingest. but I think, Given that we know the Category, you know. It's a little bit. Assumption on outside then that you say Hey this is a an outflow. even if all the numbers are positive, you know, there will be negative in palm

54:40 Federico Morando: Yep. See your point.

54:43 Emma Sjöström: It's more about. What I really want to do is we I want us to build something that's kind of easy to use. That doesn't require too many changes americansparent with you now, but on the customer side of things,

54:56 Federico Morando: Okay.

54:58 Emma Sjöström: Um, certain types of numbers formatting might be a bit unconventional, but if this is for example, this one is something that you want to use. Will make it happen. and I think it's very useful for us regardless to see this different ways data can be formatted because we need to make sure we can parse what you

55:19 Amanda Mitt: Yeah. But this is like is a surprise even for us this format. Um, we don't really use it, but the thing is we're not owners of this Google sheet, so that's why we're so reluctant on asking changes.

55:32 Emma Sjöström: Yeah, I know. It totally understand. so,

55:36 Federico Morando: Yeah, or like even just would be good to from now on say. Yeah because they are dated. These are details that Don't really make that much of a difference like it's like, it's just about the amount. And what's the formatting? So, Like just to know that from for example, from now on if it's better for you to recognize it in a proper standard format, we can have it like that and sorry it's just a matter of alignment with the teams and things.

56:09 Federico Morando: So

56:14 Emma Sjöström: Yep, that's also fair.

56:16 Federico Morando: Okay.

56:17 Emma Sjöström: Let's check how big of an issue it is to support this type of formatting as well. And get you. And also look into what happened there. On the column side in the mappings. And It seems, it could have been a UI issue that it wasn't possible to scroll all the way down or fetch, the right one.

56:38 Emma Sjöström: It could also be an issue with us not being able to manage. Above a certain number of columns.

56:47 Amanda Mitt: Yeah. Also, I think I don't really feel like I would really think about. For example, if he's doing this weekly, we're here, upload the full file? Or would he check just the upcoming, you know? Payments and then just add that and like delete the previous rows um just thinking about it.

57:04 Amanda Mitt: Like I think the easiest way would be performed to recognize that it's a fast forecast underneath it.

57:11 Federico Morando: As with.

57:12 Emma Sjöström: And so you it's a focus.

57:14 Federico Morando: Sorry, I think I think I lost you. Sorry.

57:17 Emma Sjöström: oh,

57:19 Amanda Mitt: Because for example, when he you're, let's imagine this, everything is working

57:19 Federico Morando: The.

57:23 Amanda Mitt: and you're doing the process weekly. You're you want to include the tax payment forecast, right? Are you should we think about him just downloading the file and uploading everything? Even the past, or would he do it? Ideally do a check and just upload the future values, right? The future payments.

00:00 :

57:45 Emma Sjöström: So, how problem works is? We would only allow ingesting future dates For forecast, we will only say, You know, it will end up as an invalid role. It will be like, Oh, did they asked? So we wouldn't be included.

57:56 Amanda Mitt: Okay.

58:00 Emma Sjöström: And we also have this. Now we didn't really get a good chance to see that but we have this kind of duplicates detection feature as well. So if you were to if there's no other changes on a row We will say, Hey, you've already added this one. If it's the same date, the same everything so that should cover it.

00:00 :

58:20 Amanda Mitt: Okay. Then we think about yeah like we just need to. Yeah think about the

58:21 Emma Sjöström: And we,

58:25 Amanda Mitt: process like that that we will always check delete and upload. Just

58:29 Emma Sjöström: Yeah.

58:30 Amanda Mitt: New new payments, okay.

58:31 Emma Sjöström: My ambition here is actually to make it easy for you guys. I know that makes

58:35 Amanda Mitt: Okay.

58:36 Emma Sjöström: messages here, but to me this is very valuable because I want the data that lives in files ultimately should be easy for you to upload. Because while there might be some special cases where it's really helpful. If you, you can amend your internal processes and align with other teams, It would be amazing if we can make this feature happen in a way.

59:03 Emma Sjöström: That's like, not Creating more friction for you guys, just

59:07 Federico Morando: Yes.

59:08 Amanda Mitt: Yeah.

59:09 Emma Sjöström: so actually,

59:09 Amanda Mitt: But we can also be flexible. And this change, the wouldn't be that hard. Honestly would be just one time a week would be fine. Yeah.

59:14 Emma Sjöström: Yeah. Yeah.

59:18 Federico Morando: Yeah, since a lot like it's not a tons of like payments like we can adjust to

59:23 Amanda Mitt: Yeah.

59:24 Federico Morando: that obviously. So

59:26 Emma Sjöström: that sounds great but that being said Don't be afraid to like Do this feedback and say Oh we would prefer This is really good feedback for us and our learning also because we're building this AI, assisted file, upload ingestion flow, and it's a bit It's the V zero. It's like a very, very basic version of it up now but we need this kind of feedback to make it really good.

59:49 Emma Sjöström: so,

59:50 Amanda Mitt: Okay.

59:51 Federico Morando: Yeah.

59:51 Amanda Mitt: Yeah.

59:52 Emma Sjöström: And that being said, we want to make sure it works for you as soon as possible and then we can keep iterating. But this is super good. Super helpful feedback.

59:58 Amanda Mitt: Okay.

01:00:00 Federico Morando: Thank you, we appreciate from your side.

01:00:01 Amanda Mitt: Thank you.

01:00:02 Emma Sjöström: Thank you. Okay, I have to go, I have another meeting, but this was super helpful and I feel that we will definitely dive into this again. But

01:00:13 Federico Morando: Yes.

01:00:15 Emma Sjöström: Yeah.

01:00:16 Amanda Mitt: Okay.

01:00:17 Emma Sjöström: Okay.

01:00:18 Amanda Mitt: Thank you very much.

01:00:19 Emma Sjöström: Thank you.

01:00:20 Federico Morando: thank you so much, it was Right.

01:00:26 Amanda Mitt: Bye.

01:00:27 Emma Sjöström: but,

01:00:27 Federico Morando: Let's see. Bye.