Partner-driven lending. Connected operations.
Connected lending starts by following the money.
Lending is becoming partner-driven. Money flows are more complex. AI is changing the work. Regulation is getting tighter.
Software built around separate LOS and LMS silos misses the workflows between them. We are building a connected lending stack, starting after approval.
The system needs to connect the loan, its partners and every movement of money.
Partners and money flows
The operating model is changing
Lending is becoming partner-driven. Money flows are becoming more complex.
Lending increasingly depends on sourcing partners, servicing partners, lending service providers and co-lenders. Each participant has a different role in the loan’s life.
The money flow depends on the product and the partner agreement. Payments need to be allocated. Partners need to be paid. Settlements need to be reconciled. Exceptions need an owner.
Trace partner payouts to the agreement and the loan. Funds move through lender accounts and authorised banking rails.
Customer access
The partner network is a moat
Customer access through sourcing and servicing partners is a moat.
Sourcing partners open access to customer segments a lender may not reach on its own. Servicing partners bring the relationships and local capability needed to support those customers after approval.
That access is a moat when the lender can make the network work: onboard partners, define responsibilities, track the work and settle payouts correctly.
Customer access depends on the network. The operating system needs to support the work behind it.
AI in loan servicing
Loan servicing is next
AI will reshape lending products, operations and customer experience.
We are seeing AI’s impact mostly in sourcing today. Next, it will change loan servicing: how teams follow up, how borrowers get help and how lenders evaluate their policies.
- 01 / Proposal
AI proposes the work.
Use the loan’s context to propose a servicing action or a policy change.
- 02 / Authority
The core controls execution.
Apply lender policy, permissions and named approvals before money or loan state changes.
- 03 / Evidence
The record keeps the evidence.
Connect the proposal, approval, action and outcome to the same loan.
AI belongs in the workflow.
The deterministic core executes authorised actions. The lender remains accountable.
Regulation and governance
Controls need to work at the point of action
Regulation is evolving, and controls are getting tighter.
As partners and AI take on more work, the lender needs clear controls over who can act, which policy applies and what evidence is kept.
Define the policy
Set permissions, limits and the approvals an action requires.
Enforce the controls
Check the action against the approved policy. Hold it when the conditions are not met.
Preserve the evidence
Keep the actor, policy version, approval and outcome available for review.
Compliance cannot depend on reconstructing events from separate systems. The controls and the evidence need to be part of the operating workflow.
A connected lending stack
Follow the money trail
LOS and LMS silos miss the workflows that connect the loan.
The industry has been poorly served by products that split lending into LOS and LMS silos. Too many workflows fall between them: partner onboarding, payouts, servicing, collections, recovery and reconciliation.
Those products were not built around partner-led operating models or AI in the workflow. Adding another tool leaves the same gaps between the loan, the people doing the work and the money moving through it.
We are building a connected lending stack by following the money trail. Our focus today is post-approval.
- Participants
- Product and policy rules
- Money movements and evidence
The lender’s core or LMS remains the authoritative system of record for the chosen deployment.
Our focus today
The connected stack after approval.
The LMS is one part of the work. Lokta connects the workflows around the live loan book.
- 01
Loan Management
The LMS for the live loan book.
- 02
Loan Product Studio
Product rules that carry through the loan’s operating life.
- 03
Partner onboarding and payouts
LSP to NBFC. NBFC to LSP. Responsibilities and payouts tied to the product and partner agreement.
- 04
Collections
Follow-up and collection workflows under lender policy.
- 05
Borrower servicing
Requests, repayment context and servicing actions on the same record.
- 06
Recovery
Recovery work connected to the loan and its history.
- 07
Reporting
Reports drawn from the operating record.
- 08
Compliance
Controls and evidence inside the workflow.
- 09
Analytics
A view across loans, partners and operating outcomes.
Loan Origination is on the roadmap. Our current focus starts after approval.
The platform foundations
Built for partners, AI and connected workflows.
Four foundations carry the operating model.
- 01
API-first
Connect lender systems, partner systems and workflows through APIs.
- 02
Governed AI
AI proposes actions. Policy checks and required approvals control execution. The record keeps the evidence.
- 03
A strong data backbone
The lending ontology connects loans, participants, money movements, policies and outcomes.
Explore the lending ontology - 04
AI-led policy governance
AI helps evaluate and propose policy changes. The lender approves them, and the deterministic core enforces the approved policy.
Why Lokta
Built from experience. Designed for the next operating model.
We are the team behind Apache Fineract. We later built and scaled a commercial lending platform on those foundations. Lokta is what we would build if we started today.
Loan Management, AI Loan Servicing and RBI Model Risk Management are available now. Loan Origination is on the roadmap.
The detail
The questions behind the thesis.
Why does partner-driven lending change the software requirements?
Each product and partner agreement changes how money is allocated, settled and reconciled. The system needs to connect those movements to the loan, the participants and the agreed rules.
What does following the money trail mean?
Trace each payment, allocation, partner payout, settlement and recovery to the loan and the agreement that governs it. When a movement creates an exception, its owner and resolution should stay on the same record. Funds move through lender accounts and authorised banking rails.
Where is Lokta focused today?
Post-approval: Loan Management, Loan Product Studio, partner onboarding and payouts, collections, borrower servicing, recovery, reporting, compliance and analytics. Loan Management, AI Loan Servicing and RBI Model Risk Management are available now. Loan Origination is on the roadmap.
How does Lokta govern AI and policy changes?
AI proposes servicing actions and policy changes. Lender policy, permissions and named approvals govern what can proceed. The deterministic core executes authorised actions and preserves their evidence. The lender remains accountable.
Does a connected lending stack include origination today?
Our connected lending thesis spans the loan’s life. Our current focus is after approval. Loan Origination is on the roadmap and is not available now. The hand-off from origination needs agreed data, reconciliation and an owner for exceptions.
A thesis from building lending systems.
This is the operating model we are building for: partner-driven lending, governed AI and a connected record of the loan and its money flows.