ONESOURCE DataFlow
Collect data from every entity once, and report across all of it.
Every year the same request goes round, asking the legal entities for data that no system can give you. Weeks later the last reply arrives, and somebody opens every attachment by hand, moves the data somewhere it can be used, and starts the analysis. ONESOURCE DataFlow is a different way to do it: corporate tax data collection from every group entity, through one standard template.
The analysis is not the problem. A qualified person should be doing that. It is everything around it: the chasing, the reformatting, the rekeying, and next year the same again with none of last year's decisions written down anywhere you can reach.
Thomson Reuters describe ONESOURCE DataFlow as tax data management software: standardise your data collection packages and workpapers, add controls and validation, roll them forward each year, and flow the data into a central database you can report on. All true, and very underused. Two things that description leaves out.
Thomson Reuters ship the engine. You define the workbook, so the process is the one your business actually runs, and it can keep changing as you do.
Which is where finance people already work, and which means you can build automation on top of it.
Collect, classify, connect, learn
DataFlow is the focus. The walkthrough is one example of what can be built on it, and of what AI and APIs can add once it is in place.
Collect
The request pulls its own trial balance and takes the detail, pasted in or read from documents.
Classify
Rules run first, then AI proposes a category, a confidence and a reason for every line.
Connect
One button sends the result into the computation through an API, with nothing rekeyed.
Learn
Every line a reviewer overruled is reported, and can become a rule for next year.
The AI and the APIs are what you add once the collection is right. They are not the reason to start.
Is a DataFlow collection worth building?
Ask these of any process you run. A yes to any of them means it is worth a conversation.
- 1Is the same information collected from several people, more than once?
- 2To report across all of it, do you have to open every file?
- 3Is more time spent managing the process than doing the work?
- 4Does each entity end up with its own version, and next year a new one?
And one that rules it out. If a system can already give you the data, this is the wrong tool. That is a report, not a collection, and we would tell you so.
Where else DataFlow works
Tax
- Capital allowances
- Transfer pricing balances
- R&D project costs
- Lease data
Finance
- Consolidation packs
- Intercompany matching
- Statutory disclosures
Anything a team currently emails a spreadsheet for.
ONESOURCE DataFlow: frequently asked questions
Are you a ONESOURCE DataFlow consultant and implementation partner?
Yes. Osprey Consulting is an authorised Thomson Reuters ONESOURCE implementation partner and a ONESOURCE DataFlow Certified Implementer. A certified implementer works on every implementation. You work directly with Mark and Charlotte throughout, with no handoffs to junior consultants.
Does the AI make the decision?
No, and AI is not the focus of what we offer. The focus is DataFlow and what it gives you: one standard template for every entity, validation and controls, a roll forward each year, and reporting across all of it from one place. The walkthrough is one example of what AI and APIs can add within that framework. How AI behaves in any DataFlow or wider Excel implementation is governed by the code written for it. In this example rules run first, then the AI proposes a category, a confidence and a reason for every line, and a reviewer overrules it where they disagree. Every overruled line is reported.
Do we need General Ledger Manager?
No. In the walkthrough the trial balance arrives from General Ledger Manager, but it can come from another source, and if it fails to arrive the workbook says so visibly. The example shows what DataFlow can do, and it is only one type of implementation. Contact us to discuss yours.
Which systems can the results flow into?
The walkthrough shows ONESOURCE Corporate Tax. Because it is an API, the same approach works with any API enabled system.
Does DataFlow work with Alteryx?
Yes. DataFlow runs the collection, and Alteryx can handle the data work around it: preparing what goes in, and moving the results on to other systems. We are a Registered Alteryx Solution Provider Partner, so we can build both sides.
Is it only for tax?
No. It suits anything a team currently emails a spreadsheet for, including consolidation packs, intercompany matching and statutory disclosures.
Who maintains the workbook afterwards?
Your team. The template is yours, and documentation and training are part of the work, so your team maintains the workbook after we have gone.
Do we need a ONESOURCE DataFlow licence?
Yes. To use ONESOURCE DataFlow you need to license it from Thomson Reuters. Get in touch and we can talk through what you need in more detail, and help you work with Thomson Reuters on it.
Does this work with DataFlow Classic?
Our example is built on DataFlow, not DataFlow Classic, and DataFlow is what we implement. Classic has some limitations around the use of macros, which the AI in our example relies on.
Where does our data go when the AI classifies it?
To whichever AI you choose to use. Our example uses Grok, but any AI could be used, and that is the system that would consume your data. It is a choice you make as part of the design, so it should be an AI your organisation is comfortable sending that data to.
Considering ONESOURCE DataFlow?
Let's have a straightforward conversation about your situation - no lengthy proposal, no hard sell.