We in the DataPal team have been developing our core product for the last 12 months, and actively running our ’Simulator’ environment for 9. We’ve learned a lot along the way. In this post we summarise those learnings, and begin to point the way forward to the live product launch. Many of these themes we’ll come back to and drill into the detail over the next few weeks and months.

What have we learned:

1. A smart digital filing cabinet run for people on a fiduciary basis supports many useful functions, because more and more of the records of our lives and the things we have are now digital.
2. Just as it makes perfect sense on many dimensions for organisations to maintain records of their customers, the same is the case for individuals maintaining records of the organisations they deal with. But from the individual perspective this is more of a tactical accumulation over time than a conscious plan.
3. People record, typically, 60-80 items in their DataPal accounts; and 30-40 organisations or apps they have would have connections with if possible.
4. Most prefer to have web/browser access for data entry, and a mobile app for portability and accessibility.
5. The highest impact data feed is the open banking one. Most people make/ingest 5-10 transaction records per day.
6. But open banking records typically don’t become useful until augmented; but when that is done, they are high value.
7. The benefits that people feel they get from the approach are:
   ‍
   a) They feel more in control of their data
   b) Having direct access to their data means some processes become easier/faster
   c) Automatically generated ‘jobs to be done’ records and then automated reminders help forward planning

The types of people most likely to use DataPal in the early days are:

1. People who like to be ‘organised’ (c. 15% of our sample)
2. People who are very aware of data privacy issues (c. 35% of our sample)
3. People already working in ‘data’ related jobs or roles (c. 30% of our sample)

The benefits that organisations and app developers feel they get from the approach are:

1. Data quality improvements
2. Contract based data exchange feels more sustainable and more trustworthy than consent (with the associated un-readable privacy policies)

Organisations most likely to connect to DataPal in the early days are:

1. App developers, because they can ‘outsource’ the painful personal data related aspects of building apps to the individual/DataPal
2. Existing data service providers (e.g. verification services) who now have another revenue stream
3. Organisations using Salesforce CRM/AgentForce (that is because of our teams deep knowledge of those platforms, and the connections to them built in our Simulator).

The main thing that has changed through the Simulator period is the rise of AI Agents as a technical capability and market-facing proposition to figure out how to best deploy. Critically there is now a very clear distinction between ‘fiduciary’ agents (those working contractually for the individual), and non-fiduciary (those working for other parties). Our research suggests that it is incredibly difficult to build the trust that agents need to operate when their paymaster is someone else.

Then another major innovation emerged that caused us to change our deployment plans. FedID, and FedID Connect (a safer version of OpenID Connect from the individual perspective) was conceived and open-sourced by our technology partners at JLINC Labs. For DataPal that adds greater security and trustworthiness than is available from other ‘sign up/sign in’ methods, with also more downstream utility.

All of that means that our DataPal proposition is now very much around Personal Fiduciary AI powered agents working on behalf on individuals to help them get more done, without any loss of control, and building value from their data over time.

This short video and visual sum up where are headed.

‍

‍

‍