Finance training through realistic banking decisions: decide, see the consequence, learn why.

How a decision is set up

1. A case, not a question: a fictional client or market situation with the papers a banker would see, such as a borrower’s accounts, a client’s exposure, the bank’s limits and its policy.
2. The figures are the bank’s: DSCR, debt to equity, forward rates, the LCR and the rest are worked out by the platform’s calculators, so the learner spends the time on judgement, not arithmetic.
3. A choice with a reason: the learner picks a course of action and writes a short rationale, which is part of the score.
4. A consequence: the next round shows what the decision did, and some cases come back later, as when a loan is checked again twelve months on.

Scored on policy and process, not hindsight

Each decision is scored by fixed rules, not by AI: the analysis, the risk taken, the decision itself, the written reason and whether the learner stayed inside the limits. Where a case states a policy, the decision is marked against that policy. An FX hedge, for example, is judged by whether it met the treasury policy in the case, never by where the market happened to go afterwards. A hedge that looks expensive in hindsight can still have been the right call in advance, and learners are taught to tell the two apart.

Explained, so the lesson sticks

After each result Trezi and Professor Owl explain what happened and why, in plain English or Hindi. Professor Owl’s classes, with Riya and Kabir, teach the idea before the desk. The AI never makes up a number: it explains figures the bank has already calculated, and a person approves every action.

Decisions with a second pair of eyes

Banks do not let one person enter and approve the same thing, and neither does TrezCap. With maker-checker, one learner enters a deal, a payment or an account and another checks and approves it, so a class practises catching errors as well as making decisions. Some checker exercises carry planted mistakes to find.

Four decisions to start with

• The 100 Crore Credit Decision: how much to lend, at what rate and on what covenants, then a credit committee and the loan twelve months later.
• The FX Dealing Desk: book a client’s forward, take the market move and decide how much to cover inside the limit.
• The Liquidity Crisis Morning: deposits run off and a credit line is pulled; keep the LCR above the floor without a fire sale.
• The Governed AI Bank: AI agents draft the day’s work; you set their limits, approve or stop them and handle an incident.

For a class or a batch, faculty and trainers see every learner’s scores and can run a debrief after a simulation.

Free for now.