AI-native lending software, not a chatbot bolted on
Most lending software has added an AI panel this year. The useful question is narrower: which job that you currently do by hand does it actually take off you? Here are the three Byaj Batta does, each against your live book rather than an export.
Tezu — ask your book a question
Tezu is the agent inside the product. Ask what is overdue past thirty days, which branch slipped this month, or how a particular borrower has repaid, and it answers from your own data. The value is not the conversation — it is that the answer needed no report to be built, no filter to be found and nobody to be asked.
AI import — arrive without retyping
The reason lenders stay on a spreadsheet is rarely the spreadsheet. It is the migration. AI import reads the file you already keep — whatever the columns are called, however inconsistently — and proposes a mapping onto customers, loans, groups, deposits or shareholders, with a transform per column. You review and correct the mapping before a single row is written, so the AI does the tedious matching and you keep the decision.
Insights — what changed, without asking
Portfolio insights surface movement worth noticing: collection efficiency drifting, overdue concentrating in one officer's book, a branch pulling away from the rest. The point of putting this in the dashboard rather than a report is that nobody has to remember to look.
Where your data goes
Your portfolio data is used to answer your questions and to map your imports. It is not used to train models. Each organisation's data is held in its own database schema, so one lender's book is never queried alongside another's.
Frequently asked
What is Tezu?
Tezu is the AI agent built into Byaj Batta. It answers questions about your own portfolio in plain language — what is overdue, which branch is behind, how a borrower has repaid — without you building a report first.
Can AI import my existing data?
Yes. AI import reads the spreadsheet you already keep and maps its column headings onto customers, loans, groups, deposits and shareholders, proposing a mapping and a per-column transform which you review before anything is written.
Do I have to rename my columns before importing?
No. That is the point of the AI step. It matches whatever the columns are called in your file, including inconsistent or abbreviated headings, and shows you the mapping it proposes so you can correct it.
Is my portfolio data used to train an AI model?
No. Your data is used to answer your own questions and to map your own import. It is not used to train models.
What does AI-native mean here?
That the AI is part of how the product works rather than a chat box added on top — it reads your book to answer questions, does the mapping work during migration, and surfaces portfolio insights, all against live data rather than an export.