About · History

From Apache Fineract to the agentic loan servicing platform

When you pick a loan platform, you also pick the people who will maintain it for years. This page dates what this team built before Lokta, and in what order, so you can check the record against the claim.

  1. Part one

    Mifos and Apache Fineract

  2. The Mifos software effort begins

    Mifos’s own history dates the start of the software work to this year.

  3. Mifos is released as open source

    The code is published openly.

  4. Ashok Auty co-creates Mifos X and Apache Fineract

    Ashok, now a Lokta co-founder, is one of the people who shapes the lending core and the Mifos X distribution built on it.

  5. The platform enters the Apache Incubator

    The code is donated to the Apache Software Foundation.

  6. Apache Fineract becomes an Apache Top-Level Project

    From here it is governed as an independent Apache project.

  7. Fineract is recertified as a UN Digital Public Good

    Apache records the recertification by the Digital Public Goods Alliance. This is the project’s milestone; Lokta is not affiliated with the Fineract or Mifos communities.

  8. The ecosystem reaches 65M+ borrowers in 70 countries¹

    Mifos reports 500+ financial institutions using the ecosystem. Lokta’s own central estimate of the loan principal disbursed through it, from 2006 to early 2026, is about $500B.

  9. Part two

    A commercial lending platform on those foundations

  10. The same team builds and scales a commercial lending platform

    The team extends the open-source foundation into a commercial product for Indian NBFCs and fintechs: more configurable than the open core, with the operations and support a regulated lender needs. It works.

  11. The limits turn out to be architectural

    Overnight batches make lenders wait on yesterday to act today, and a data model per module leaves someone reconciling spreadsheets at month-end. The team concludes these limits live in the architecture, so no upgrade removes them.

    Ashok Auty on the three lessons

  12. Part three

    Lokta

  13. Lokta starts on a new codebase

    The team starts over with a deterministic core for money movement, one shared ontology of the loan, and AI that works inside governed workflows. Lokta is written agent-native from the ground up, and it does not fork the Fineract code.

  14. Lokta publishes its estimate of the Fineract ecosystem

    The post gives the full range behind the $500B figure and the assumptions it rests on.

    The estimate and its method

  15. Ashok Auty writes down why the team started over

    The post pairs each Fineract lesson with the design choice Lokta made in answer.

    What we learned building Fineract

  16. The core ships as seventeen modules in one binary

    Each lending domain is its own Gradle module with enforced boundaries, and all seventeen deploy as a single Spring Boot application, so a disbursement commits or rolls back as one transaction.

    Why Lokta is polylithic

  17. Chandramouli sets out the case for Lokta

    Lokta’s co-founder and CEO, a friend of Ashok’s for 26 years since college, explains why the live book after approval is where the work is.

    Credit is the bloodline of the economy

  18. The five governance principles are published

    Lokta states how agents act on a live book, and opens work with a small set of co-design partners for 2026: lenders, fintechs and LSPs.

    The principles

  19. Where the platform stands now

    Loan Management, AI Loan Servicing and RBI Model Risk Management are available now. Loan Origination is on the roadmap.

    The platform today

¹These are totals for the whole Mifos and Apache Fineract ecosystem the team helped build. Lokta does not count them as its own book, and the $500B is a modelled central estimate.

Talk to the team behind this record

Lokta is founder-led. If you are weighing it for a live book, the people in this timeline are the people you will speak to.

Founder-led adoption

Adopt the agentic loan servicing platform.

Lokta is built for enterprise deployment, VPC or single-tenant cloud, with an audit trail in every state change. We work with a select group of institutions through a founder-led model: deep adoption, deliberate scope, a delivery window the team commits to in writing.