Odysı
See if we fit
Solutions
Solution Fintech & financial literacy

AI for financial literacy in fintech

For a fintech, education is not a side project. People adopt savings, investing, credit, and insurance only when they understand them, so in-app education is a direct activation lever. The problem is that most of it never gets finished.

O
Odysi Solutions
Proof Overview Effect, built for GCash
Read 6 min

Most in-app education lives as long, help-center style articles that are formal, buried, and rarely finished. Getting it to actually land is partly a product problem and partly a place where AI can help, as long as it is pointed at learning rather than time spent. This is a practical look at why education is an activation lever, where AI genuinely helps, and how we approached it inside a large mobile wallet.

01

Why in-app education is an activation lever

When a platform serves many people who are new to formal finance, the growth that matters is helping transactional users take the next step into products like savings or credit. Those products get adopted when people understand them, which makes education a lever on activation and not a compliance checkbox. Treating it that way changes what you build and how you measure it.

The education that already exists in most apps is written to inform, not to be finished. It is long, formal, and disconnected from the moments where someone actually makes a money decision. The few who start tend to leave before the end, and there is no feedback, no sense of progress, and no link between a lesson and the action it should unlock.

The article nobody finished drop-off
Measured by time spent
A path people finish completed
Measured by learning completed
Fig. 1: a buried article rewards time spent and loses people line by line. A short path with progress and feedback is built to be finished.
02

What works, and where AI helps

The design principle that matters is simple: measure learning completed, not time spent. A game-like experience that people finish, with clear progress and feedback, tied to the moment a decision is made, does the activation job that a buried article cannot. Get that right first, because no amount of AI fixes education that is built to be long rather than to be finished.

Within that design, AI extends what is possible:

Personalisation

Adapting the path to someone's level and the product they are closest to adopting, so the lesson is relevant.

Adaptive pacing

Adjusting difficulty as someone learns, so they are neither bored nor lost.

Contextual delivery

Meeting people at the moment of a money decision with the right short lesson.

Content at scale

Generating and tailoring variations so material stays fresh, under human review.

The goal is that people learn and act, not that they spend more time in the app. Optimise for learning completed and real decisions made, and keep the design free of dark patterns.

03 · Product & design lead

What we did for Overview Effect

Overview Effect brought us in to lead product on a financial-literacy learning experience inside GCash, the dominant mobile wallet in the Philippines. Our remit was to design a game-like system people would actually finish, built on one rule: measure learning completed rather than time spent. It replaced long, buried articles that few people opened and fewer completed.

Read the Overview Effect case study
04

How to start without overcommitting

Pick one product you want people to adopt, and design the shortest game-like path that genuinely teaches the decision behind it, measured on completion and on the action it unlocks. Add personalisation and contextual delivery once the core experience is finishable. The aim is narrow: turn education from a buried article into a lever that moves people to the next product, without resorting to engagement tricks.

Common questions

FAQ: AI for financial literacy

Can AI teach financial literacy on its own?
AI helps personalise, adapt, and deliver learning in context, and it can generate material under review. The experience still has to be designed to be finished and tied to real decisions; AI extends a good design, it does not replace one.
Why do in-app financial education efforts usually fail?
Because they are built to inform rather than to be finished: long, formal, buried, disconnected from the moment of decision, and measured by time spent rather than learning completed.
How should you measure a financial-literacy feature?
By learning completed and by the financial actions it unlocks, not by time spent or pageviews. Time-based metrics reward the wrong behaviour.
Does personalisation with AI risk manipulating users?
It can, if it is aimed at engagement. The safeguard is to optimise for learning and real decisions, not time in app, and to avoid dark patterns by design.
Is this only for large wallets like GCash?
No. Any fintech that wants users to adopt savings, credit, insurance, or investing can treat education as an activation lever. The scale changes the numbers, not the approach.
Keep reading
Free field assessment Is your AI project worth building? Take the AI Project Scorecard. Five questions, two minutes, an honest verdict: build, fix, or wait. Take the test →
Prototype. Automate. Grow.

Education that nobody finishes?

If your in-app education is a set of articles few people finish, and you want it to move users into your products, that is a product problem worth solving, with AI extending the design rather than replacing it. We are easy to talk to.