← Agent catalogue

AI Agent Profile · LendingIQ · Bengaluru

Chief Compliance Officer AI

Invoked via: internal orchestration APIRuntime: AWS Bedrock · ap-south-1Model: Claude Sonnet 4Context window: 200K tokens

DivisionCompliance

Resume

What this agent does

The CCO AI monitors the regulatory environment for changes that affect LendingIQ, interprets what those changes require, drafts the internal policy and board reporting responses, and prepares the compliance intelligence that a human CCO needs to discharge their regulatory duties. It is the research, drafting, and monitoring engine of the compliance function — not a replacement for the human accountability that RBI regulations require to be named and present.

Primary functions

Regulatory Change Radar

Triggered on RBI / MCA / SEBI publication

Invoked when: new circular, press release, or amendment detected in the regulatory corpus pipeline

  • Monitors the RBI website corpus pipeline — Master Directions, circulars, press releases, FAQs, and amendment notifications — and triggers an alert when new content is detected that is relevant to an NBFC operating in LendingIQ's product segments.
  • Reads the new circular in full and produces an impact summary: what the regulation requires, which specific LendingIQ operations are affected, what the compliance deadline is, and whether any existing internal policy is now non-compliant.
  • Cross-references the new requirement against every existing internal policy document in the RAG store to surface the exact clauses that need amendment — not a generic "review your policies" notice but a specific diff showing where the gap is.
  • Does not interpret regulatory intent where the circular is ambiguous. It flags ambiguity explicitly and identifies the specific provision that is unclear, so the human CCO can seek informal RBI guidance if required.
Output: Regulatory change alert — circular summary, applicability assessment for LendingIQ, compliance deadline, list of internal policies requiring amendment with specific gap clauses identified, and ambiguity flags requiring human judgement.

Internal Policy Formation

Triggered on regulatory change alert or internal review cycle

Invoked when: regulatory change requires policy update, or annual policy review cycle due

  • Reads the triggering regulation alongside the current internal policy (both via RAG) and drafts the specific amendments needed — in the same language register, clause numbering, and document structure as the existing policy so the committee sees a clean redlined document, not a rewrite.
  • Where a new regulation introduces a requirement that no existing policy covers, drafts the new policy section from scratch — grounded in the regulatory text, referenced to the specific RBI clause, and consistent with the broader policy framework.
  • Checks every drafted clause against the full regulatory corpus for internal consistency — a new AML clause must not conflict with an existing KYC section, a new FLDG policy must not contradict the existing co-lending guidelines.
  • Cannot determine what the organisation chooses to do where the regulation provides discretion. Where RBI gives NBFCs a choice — e.g., whether to adopt the simplified KYC route — the agent drafts options with implications of each; the human compliance committee makes the election.
Output: Redlined policy document — clause-level amendments with regulatory citations, new sections where required, internal consistency check results, and a summary of discretionary choices awaiting human decision.

Board Reporting

Triggered on quarterly board cycle or ad-hoc request

Invoked when: quarterly board pack due, board committee meeting scheduled, or specific compliance query raised by a director

  • Reads the compliance calendar, open regulatory action items, internal audit observations, and any new regulatory changes issued in the period — all injected at invocation — and synthesises them into a board-ready compliance report.
  • Structures the report to what a board needs to know and decide: regulatory changes and their impact on the business, compliance status against key obligations, open action items with owner and deadline, and items requiring board approval or resolution.
  • Drafts the compliance committee minutes summary, the compliance attestation language for the board pack, and any board resolutions required for regulatory purposes — all for human review, approval, and signature before use.
  • Does not certify compliance on the board's behalf. The board resolution and sign-off remain human acts. The agent prepares the materials and the draft language; the directors deliberate and approve.
Output: Board compliance report — regulatory change log, compliance status dashboard, open action item tracker, items for board decision, draft resolutions, and attestation language. All marked "draft for board review" until approved by the human CCO.

RBI Relationship Management — Preparation

Triggered ahead of inspection, submission, or meeting

Invoked when: RBI inspection scheduled, regulatory return due, or RBI query or notice received

  • For inspections: reads the last RBI inspection report and action-taken report (ATR), identifies which prior observations have been closed versus open, and prepares the compliance readiness brief — what the inspection team is likely to focus on, what evidence needs to be assembled, and what residual gaps the human CCO should be prepared to address.
  • For regulatory returns: reads the prescribed format, maps each data field to the internal data source, flags fields where data may be incomplete or require manual verification, and drafts the narrative sections — all for the human CCO to review and submit.
  • For RBI queries or show-cause notices: reads the notice, identifies the specific regulatory provision cited, and prepares a draft response — factual, measured, and grounded in the regulatory text. The human CCO reviews, modifies, and sends the response under their own signature.
  • Does not communicate directly with RBI. It has no authorisation to send correspondence, attend meetings, or speak on behalf of LendingIQ with the regulator. All communications go through the named human compliance officer.
Output: Inspection readiness brief, pre-populated regulatory return drafts, or draft responses to RBI queries — all clearly marked as draft, prepared for human CCO review and despatch.

Risk Appetite Framing

Triggered at annual strategy cycle or regulatory change

Invoked when: annual Risk Appetite Statement (RAS) review due, or material regulatory or portfolio change warrants reassessment

  • Reads the current Risk Appetite Statement, the latest portfolio performance data, the regulatory framework constraints, and the board's stated strategic objectives — and identifies where the existing risk appetite parameters are out of step with any of these inputs.
  • Drafts proposed revisions to the RAS — updated tolerance thresholds for credit risk, operational risk, compliance risk, and liquidity risk — with the rationale for each change and the regulatory minimum that sets the floor below which the appetite cannot be set.
  • Coordinates the RAS framing with the CRO AI's risk assessment output — the risk appetite must be set above the floor of what the portfolio can actually sustain, not just what the board would prefer. Flags misalignments between stated appetite and observed portfolio risk for human resolution.
  • Cannot determine what risk appetite the board should choose — that is a governance decision that belongs to the board and reflects their commercial judgment, capital position, and stakeholder obligations. The agent provides the analysis and drafts the options; the board decides and owns the statement.
Output: Risk Appetite Statement revision draft — current vs proposed parameters, rationale for each change, regulatory floor citations, misalignment flags between appetite and observed portfolio risk, and a board discussion note on the tradeoffs.

Knowledge base

RBI Regulatory Corpus (live RAG)

Master Directions for NBFCs, all circulars, press releases, FAQs, and amendment notifications. Ingestion pipeline must be kept current — stale regulatory corpus is the primary failure mode for this agent.

Internal Policy & SOP Library

All LendingIQ internal policies, standard operating procedures, board resolutions, and compliance manuals — retrieved via RAG at invocation time, always current version.

Prior RBI Inspection Reports & ATRs

Full history of RBI inspection observations and action-taken reports. The institutional memory of what the regulator has flagged before and how LendingIQ responded.

Compliance Calendar & Return Schedule

All regulatory submission deadlines, periodic return schedules, board certification dates. Injected at invocation to drive deadline-aware reporting and alerts.

PMLA / KYC / AML Framework

Prevention of Money Laundering Act, RBI KYC Master Direction, FIU-IND guidelines. Applied in policy formation and compliance gap checks.

General Compliance & Governance Knowledge

Pre-training knowledge of NBFC compliance frameworks, corporate governance standards, and regulatory practice from public sources up to knowledge cutoff.

Hard guardrails

Will notSubmit regulatory returns, file reports with FIU-IND, or send any correspondence to RBI. It prepares these documents; the named human compliance officer sends them under their own authority.
Will notCertify compliance to the board or to the regulator. Compliance certifications are legal attestations by named individuals. The agent drafts the language; a human signs the certificate.
Will notExercise regulatory discretion on behalf of LendingIQ. Where RBI regulations offer choices, the agent presents the options and implications — the human compliance committee makes the election and owns it.
Will notConduct or manage the RBI relationship. Regulatory relationship management involves trust, judgment, informal communication, and accountability that cannot be delegated to an AI. The human CCO holds and manages the regulator relationship.
Will notProvide a definitive legal opinion on regulatory interpretation. Its regulatory analysis is compliance intelligence, not legal advice. Where a provision carries material legal risk, a qualified legal counsel should be engaged alongside this agent's output.

Known limitations

The regulatory corpus pipeline is the single most critical dependency. If RBI issues a circular and it has not been ingested, the agent will not flag it. A missed circular that requires policy changes or a regulatory return could result in a compliance breach that the agent had no visibility of.Build a daily automated ingestion job that monitors the RBI website, MCA notifications, and SEBI circulars. Add an alert when the pipeline has not run or when ingestion fails. The agent is only as good as its inputs.
Cannot assess the tone or relationship context of RBI correspondence. An RBI query can range from routine data collection to the opening move of a serious supervisory concern — these look similar on paper but require very different responses. The agent cannot make that distinction.The human CCO must interpret the regulatory context and relationship temperature before instructing the agent on how to frame a response. Never use an agent-drafted RBI response without human judgement on the regulatory intent behind the query.
Risk appetite framing depends heavily on the quality of inputs from the CRO AI and portfolio data. If the risk picture passed to this agent is incomplete or optimistic, the drafted RAS will reflect those gaps.The RAS framing function should always be run with a joint context package from both the CCO AI and CRO AI — regulatory constraints from this agent, portfolio reality from the CRO. Neither alone gives the full picture.
Board reporting quality is constrained by what data is injected. If compliance action owners have not updated their status, or if audit observations are incomplete, the board report will reflect those gaps rather than fill them in.A human compliance manager must curate the data package before invoking the board report function — the agent synthesises and drafts from what it is given; it cannot chase action owners or validate status independently.
Knowledge cutoff applies to general compliance and governance frameworks. Regulatory practice, case precedent, and informal supervisory expectations that evolved after the training cutoff will not be known to the agent unless passed in context.Senior compliance staff with current regulatory market intelligence must review all agent outputs before external use — especially for RBI correspondence and board certifications where the stakes of being wrong are high.
Agent Profile · Chief Compliance Officer AI · LendingIQ · BengaluruLast updated April 2026 · For internal use

Important Reads

Learn more about how to deploy Chief Compliance Officer AI to your lending workflow.