← Agent catalogue
AI Agent Profile · LendingIQ · Bengaluru
Risk-Based Pricing Agent AI
Function: Pricing AnalystInvoked via: credit decision pipeline + pricing review cycleRuntime: AWS Bedrock · ap-south-1Model: Claude Sonnet 4Context window: 200K tokens
DivisionRisk division
Resume
What this agent does
The Risk-Based Pricing Agent AI ensures every loan LendingIQ approves is priced to adequately compensate for its specific risk level — computing the risk-adjusted rate for individual loans, running margin simulations for the portfolio, and monitoring competitor rate movements. It is the pricing intelligence layer between the credit risk assessment and the final rate offered to the borrower.
Primary functions
Rate-to-Risk Matching
Per approved loanINVOKED WHEN: Credit Decision Agent AI produces an approval outcome and the final rate needs to be determined
- Reads the credit risk profile of the approved application — score band, FOIR, LTV, business vintage for MSME, collateral position — and computes the risk-adjusted rate recommendation: the rate that adequately compensates for the credit risk of this specific borrower above LendingIQ's cost of funds, while remaining within the published product rate band. The recommendation is a range (floor to ceiling) rather than a single rate, because the Account Executive or Relationship Manager has discretion within the band to compete on price for high-quality borrowers.
- Applies a tiered rate structure: lower-risk borrowers (high score, low FOIR, strong collateral) receive a rate recommendation closer to the floor of the band; higher-risk borrowers (marginal score, high FOIR, no collateral) receive a rate recommendation closer to the ceiling. The risk-to-rate mapping is configured by the CFO and CRO AI — the agent applies it, it does not set it.
- Flags applications where the credit risk profile would require a rate above the published ceiling to be adequately compensated — because publishing a rate higher than the ceiling requires a rate sheet change, not an exception. The credit committee must decide whether to approve the loan at the ceiling rate (with acknowledged under-pricing of the risk) or to decline.
Output: Rate recommendation — recommended rate range (floor to ceiling for this risk tier), risk-return basis, NIM contribution at recommended rate, and flag if risk profile requires above-ceiling rate.
Margin Simulation
Monthly and on portfolio review requestINVOKED WHEN: monthly portfolio review cycle or a proposed rate change requires impact assessment
- Reads the current portfolio composition — outstanding by product, segment, and rate band — and simulates the NIM impact of proposed rate changes: if the MSME working capital floor is reduced by 50 basis points, what is the aggregate NIM impact across the current portfolio and the projected new origination? The simulation uses the current portfolio as the base and the Product Sales Manager AI's volume projections for new origination.
- Produces a margin waterfall — showing the current NIM, the impact of each proposed change, and the resulting projected NIM — so the CFO has a complete picture of the trade-off between competitive positioning (lower rate) and margin (higher rate) before approving a rate change.
Output: Margin simulation — NIM waterfall showing current and projected NIM under proposed rate change scenarios, break-even volume required to maintain target NIM if rates are reduced, and recommendation for CFO review.
Competitor Rate Watch
Weekly digestINVOKED WHEN: weekly competitor rate monitoring cycle is due
- Reads the competitor rate corpus and produces a weekly digest of significant rate movements: which competitors have changed rates on which products, the direction and magnitude of change, and the resulting competitive gap between LendingIQ's current rates and the nearest competitor for each product segment. Flags instances where a competitor rate change has created a material gap that may be affecting LendingIQ's conversion rate on high-quality borrowers.
Output: Weekly competitor rate digest — rate movements detected, competitive positioning per product segment, material gap flags, and recommended review items for the product pricing committee.
Hard guardrails
Will notPublish or communicate rates to any channel or borrower. Rate recommendations are internal inputs to the pricing process; rate publication requires CFO approval.
Will notApprove exception rates below the risk-return floor. Sub-floor rates require credit committee authority and a documented rationale for the exception.
Will notChange the cost of funds model independently. Cost of funds assumptions are set by the CFO team and are a fixed input to the pricing model.
Known limitations
Rate-to-volume elasticity is estimated, not measured precisely.The simulation of how many additional applications a 50 bps rate cut would attract is based on historical elasticity estimates and competitor volume intelligence — both of which have significant uncertainty. Treat volume projections in margin simulations as directional scenarios, not precise forecasts.
Competitor rate corpus has a lag.Competitors who change rates quietly — through DSA channels rather than published rate cards — will not be detected until the rate appears in broker comparison platforms or is reported through DSA intelligence. Supplement the corpus with weekly DSA intelligence to reduce this lag.
Important Reads
Learn more about how to deploy Risk-Based Pricing Agent AI to your lending workflow.
- Use case #0001How Pricing AI adjusts rates dynamically as risk signals changeA single interest rate applied to all borrowers is not a risk management strategy — it is a subsidy from low-risk borrowers to high-risk ones. The borrower with a CIBIL score of 790, 36 months of on-time payments, a stable salaried income, and an LTV of 62% is paying the same rate as a borrower with a CIBIL of 682, two DPD events in the last 18 months, and an LTV of 74%. The Risk-Based Pricing Agent AI ends that subsidy — and does so without manual rate-card management.Read article →
- Use case #0002Margin simulation: what Risk-Based Pricing AI models before every offerA rate offer is not a pricing decision — it is an NIM commitment. An institution that offers 9.60% to acquire a borrower without modelling the expected credit cost, the funding cost at current rates, the processing cost of this loan type, and the expected prepayment behaviour is not pricing — it is guessing. The Risk-Based Pricing Agent AI runs a margin simulation for every rate offer before it is made, confirming that the rate produces an acceptable net interest margin even at the expected loss rate for this specific borrower profile.Read article →
- Use case #0003Competitor rate monitoring: how Pricing AI tracks market rates in real timeA pricing decision made without knowing what competitors are offering at this moment — not last quarter, not last month, but today — is not a pricing decision. It is a guess dressed as a strategy. The Risk-Based Pricing Agent AI monitors competitor rates daily across all major products and geographies, alerts the pricing team when the institution's rate moves outside competitive tolerance, and models what the optimal response is before the alert reaches the desk.Read article →