I joined a five-day fintech product management cohort to learn how lenders and no-code lending platforms work. Then I planned and built DealerFinance: a loan journey for new-car buyers at a dealership. I tested it live with about 15 to 16 people and fixed what broke.
My contribution
I defined the product, wrote the PRD and sprint plan, built the journey on the no-code platform (forms, journey and logic formulas), ran the testing and wrote up what I learned.
- Banks, RBI, NBFCs and a first look at the platform
- Underwriting, KYC, credit reports, income proof, hands-on
- Office closed. Wireframe revised
- Help day. The journey takes shape
- Final brief and the live demo
- Written production launch plan
How I got in: the selection project
Before the cohort, FinStack gave candidates an assignment. I submitted a product teardown of BHIM and PhonePe, with a three-year roadmap for each. My view: PhonePe shows too much, BHIM shows too little, and each app's ownership explains why.
The page has the findings, the money table and both roadmaps, with the original PDF.
What I learned
Banks and lenders differ by what their money costs and which customers they choose. In a lending product the order of checks is a design choice: cheap checks first, costly ones later, and "no data" is a normal answer, not an error.
I wrote seven notes, and a table that links each class topic to where I used it in the product.
- What I learned, mapped to what I usedEach class topic next to where it shows up in the product, and how far it got.
- Day 1: banks, RBI and the platformHow banks, RBI and NBFCs work, and what a no-code lending platform is.
- Day 2: underwriting, KYC and credit reportsUnderwriting, KYC, credit reports, proof of income, and the platform hands-on.
What I built
DealerFinance is a six-step journey. It collects consent by one-time code, verifies identity through DigiLocker, works out the price and loan, shows which lenders are likely to accept, sends the case to one lender, reroutes after a rejection and settles the dealer's trade advance and commission.
It runs on a no-code platform, for Kia cars and four lenders. Lender terms, rates and limits are invented test values.
- DealerFinance: product requirements documentProblem, users, scope, rules and risks, in the format taught in class.
- Sprint planEpics, stories, tasks and gates.
- WireframeEvery screen with its fields, rules and test values.
What broke, and the fixes
At the live demo, customers with no credit score and customers with no car to exchange hit dead ends. I fixed both, and the second change also saved a paid lookup per case. My own testing found a loan formula that ignored the exchange value, which is still open.
The launch plan
The plan chooses a dealership system as the place to plug in, lists the fields it must give us on each screen, says what it has to change, and sets out four phases with the assumptions marked. It went in after the deadline, and the retrospective says so.
Privacy and credits
Mobile numbers, PAN, Aadhaar, emails, logins and other people's names are not in this record. The demo clip shows my own name and one-time codes on purpose and covers the rest.
Teammate: Aditya Sharan (github.com/aaDhut), Team 6.
- Consent and KYC design
- Credit bureau and soft pull
- No-code journey building
- PRD writing
- Sprint planning
- Edge-case testing
- Production launch planning
Thank you
The cohort and hackathon were an amazing experience: real products, building and experimenting across fintech. The best part was the people.
- Gaurav AgarwalCohort host, co-founder of FinStack. For walking us through how banks, loans and bonds really work.
- Dhawal PatelCohort host, co-founder of FinStack. For the platform walk-through, and the help while I built.
- Arun NandewalMentor. For connecting me with this group, and for helping organise the hackathon.
- Aditya SharanTeammate, Team 6. For being my partner in the hackathon.