What your live loan book gets on the agentic loan servicing platform
Ledgers that return the same answer every time, AI that proposes but cannot change the record on its own, and a trail your auditor can replay. Seven reasons, each with its proof.
01 A record you can recompute
Any balance on the book can be recomputed, and every change has an author.
Lokta keeps two deterministic ledgers: a Loan Ledger for schedules, balances and repayment allocation, and an Accounting Ledger for GL mappings and double-entry postings. Give them the same inputs and configured rules and they return the same figures, so an auditor can re-run a number and land on it to the paisa.
No model calculates interest or posts an entry.
02 AI that cannot act alone
AI proposes, the core decides, the record proves it, in that order.
The servicing agent drafts the borrower reply or the next action on an account. Your policy, confidence thresholds and approval rules sit in the deterministic core, and the core decides whether the proposal runs. The proposal, the approver and the resulting event all stay on the record.
The agent cannot mark a loan paid or waive a charge on its own say-so.
03 A policy that keeps improving
Your servicing policy is tested every cohort, and a person still promotes the winner.
A risk team tends to ship one policy change a quarter. Lokta is built so agents can trial hundreds of collection and servicing strategies in that time. A candidate that beats the live policy runs in shadow first, scoring real accounts beside it, and a maker-checker step decides whether it goes live.
04 Hosting you choose
You choose where the book is hosted, and the record stays inside that boundary.
The standard hosted deployment runs in-region on AWS. Enterprise lenders can run Lokta as Docker images inside their own VPC instead, on AWS or another cloud. A VPC deployment keeps the book, the models and the record inside a boundary your security team can inspect for itself.
05 A line for every partner
Sourcing partners and partner lenders each keep their own line on the loan.
Each sourcing partner has one record for diligence, effective-dated terms, attributed loans, book performance and fees. A loan holds the partner terms in force on its booking date, so a renegotiated fee does not rewrite loans already booked. Payouts are worked from those terms and pass maker-checker before they are paid.
06 The Fineract team
The people who built Apache Fineract started this platform from a blank page.
Ashok Auty, a Lokta co-founder, co-created Mifos X and Apache Fineract, the open-source lending core behind lending to 65M+ borrowers in 70 countries. The same team then built and scaled a commercial lending platform on those foundations.¹
¹The borrower and country reach is the Mifos and Apache Fineract ecosystem’s, and none of it is Lokta’s own book. How the ecosystem is sized
07 A price you can read
You can read the price before the first call.
Starter is Lokta Next 100: no platform fee for 24 months for lenders with assets up to ₹100 crore in India, or about USD 10 million elsewhere. That figure is a ceiling. A lender well below it qualifies.
What lenders said
What lenders told us on the calls.
On why Lokta starts after sanction
“The post-sanction area is the one which still today remains a grey area.”
After the portfolio dashboards
“This kind of concise and clear dashboard I never saw.”
On hearing the rate, ₹1,000 per crore of book a month
“It's a reasonable price.”
On the partner and payout view
“Traditional systems never had something like this.”
On seeing bucket roll-back and roll-forward at account level
“My team takes half a day today just to get this data out manually.”
Asked when he wanted servicing automated
“At this point, I can start within a month.”
At the end of the demo
“You have covered all the pain points of the smaller NBFC.”
Who Lokta is not for today
Lenders who need loan origination now
Loan Origination is on the roadmap. Today your own LOS, or your partner lender’s, hands each approved loan to Lokta by API.
Lenders who need a leasing module
Lokta has no leasing module.
Pure-play microfinance NBFCs applying to Lokta Next 100
The Next 100 programme, which is the Starter tier, excludes pure-play microfinance lenders.
For every other lender, Lokta covers servicing, monitoring, collections and recovery from the day a loan is approved.
Three ways to test these reasons before a call
- What lenders said when they saw Lokta
Lines from prospect calls, anonymised to role, lender type and month.
- Questions lenders ask us on calls
Integrations, pricing, migration and hosting, with every answer on the page.
- How we work with a lender
Each step from the first call to go-live, and what you bring to each one.
Reading takes less of your time than a call, but only a session on your own loans shows how the platform handles your products.
Questions about choosing Lokta
Short answers, each pointing to the page that covers the topic in depth.
How is Lokta different from a conventional loan management system?
Three things sit together on Lokta. The ledgers are deterministic, so any figure can be recomputed. Every module reads one shared model of the loan, so there is no copy to reconcile. And software agents propose servicing and collections work that the core checks against your policy before anything changes. The platform page covers the design in full.
Can Lokta’s AI change a balance or post to the ledger by itself?
No. An agent can explain an account and propose an action. The deterministic core checks the proposal against your policy and required approvals, and only a permitted action reaches the record, along with who approved it.
Do we need to replace our loan origination system to use Lokta?
No. Lokta covers the loan after approval. Loan Origination is on the roadmap, and until then your existing LOS hands each approved loan to Lokta by API.
Can Lokta be deployed in our own cloud account?
Yes. The standard hosted deployment runs in-region on AWS. On the Enterprise tier, Lokta ships as Docker images into a VPC in your own cloud account, on AWS or another cloud.
Is Lokta a version of Apache Fineract?
No. Fineract is an Apache Software Foundation project and Lokta is a separate company with its own codebase. What links them is the team: Lokta’s co-founder Ashok Auty co-created Fineract.