Sixty people, one round, and a pricing model that made every deal harder

Kaya Labs sells document-processing tools to Indian banks and insurers. It has fewer than sixty employees, a single institutional funding round, and a pricing model that its own investors initially argued against.

Kabir Sen founded the company in 2021 after eight years in product roles, latterly at Kavach Cloud. The early version of Kaya priced per seat, which is what enterprise software does, and every deal stalled at the same question: how many people would use it. The banks could not answer, because they had not deployed it. The procurement conversation could not continue without an answer.

Every deal stalled at that question, so we removed the question.

Per document

Kaya now charges per processed document. Sen describes the change as commercially inconvenient and strategically necessary. It is inconvenient because revenue becomes a function of the customer’s volume rather than their headcount budget, which makes forecasting harder and makes the company’s revenue look lumpier than a seat-based comparable. It is necessary because it removed the only question that reliably killed a deal.

The second-order effect was on the product. A per-seat vendor is rewarded for making software that many people open. A per-document vendor is rewarded for making software that fewer people need to open. Kaya’s roadmap since 2024 has been dominated by reducing the number of documents that require human review at all.

The case for staying small

Sen has said the company will not raise again until it has a second product with independent traction. The reasoning is unromantic: hiring ahead of the second product would mean spending his time managing the gap between the team’s size and its output, rather than closing it.