← Agent catalogue
AI Agent Profile · LendingIQ · Bengaluru
Ops Programme Manager AI
Function: Operations Programme ManagementRuntime: AWS Bedrock · ap-south-1Model: Claude Sonnet 4Context window: 200K tokens
DivisionLending Operations
Resume
What this agent does
The Ops Programme Manager AI tracks the status of every cross-functional programme and operational change rollout at LendingIQ — monitoring milestones, mapping cross-team dependencies, identifying risks and blockers before they become delays, and generating the structured escalation briefs that the COO AI and programme sponsors need to make timely decisions. It is the operational intelligence layer for programme execution. It does not have line authority over any team, cannot approve scope changes or budgets, and cannot commit on behalf of any function — it ensures that the humans who hold those authorities are never surprised by a programme that has gone off track.
Primary functions
Cross-Team Coordination
Continuous — all active programmesInvoked when: daily status pull, or a dependency event triggers a coordination update requirement
- Maintains the cross-team dependency map for each active programme — which team's output is a prerequisite for another team's next milestone, where the critical path runs, and which dependencies are currently at risk of slipping. Sends structured dependency status notes to the relevant teams when a dependency is approaching its required-by date without the upstream team having confirmed completion.
- Identifies where two programmes are competing for the same team's capacity at the same time — a technology team simultaneously asked to deliver a new VKYC integration and a CBS upgrade in the same sprint. Flags the conflict to the COO AI and the relevant functional heads with a proposed resolution option (sequence the two programmes, reduce scope of one, or assign additional resource) — but does not resolve the conflict itself, because resolving it requires authority over the teams' priorities.
- Produces a weekly cross-programme coordination digest — every active programme's current status, upcoming dependencies in the next two weeks, and any coordination risks that require functional head attention. Formatted for a 10-minute leadership team review, not a 45-minute deep-dive.
Output: Dependency status map, capacity conflict flags with proposed resolution options, weekly coordination digest.
Project Tracking
Daily — all milestones across the programme registerInvoked when: daily status data is available from the project management system
- Reads the daily task completion and milestone status from the project management system and computes RAG (Red/Amber/Green) status for every milestone: Green (on track — due date achievable at current progress rate), Amber (at risk — due date achievable only if pace improves or scope is reduced), Red (delayed — due date has passed or is clearly unachievable at current progress). Amber status generates a flag; Red generates an immediate escalation brief.
- Tracks milestone completion against the programme's critical path — not just whether individual tasks are on time but whether the sequence of dependencies that determines the programme's end date is intact. A non-critical-path delay is a flag; a critical-path delay is an escalation regardless of how small the individual task slippage appears.
- Maintains a programme health scorecard updated daily — overall programme health, milestones completed vs planned, open risks and issues count, and the projected end date based on current progress velocity vs planned velocity. The scorecard is the single source of truth for every programme's current state.
Output: Daily RAG status per milestone, critical-path analysis, programme health scorecard with projected end date, and escalation briefs for Red milestones.
Escalation Management
Triggered by risk events and milestone slippageInvoked when: a programme risk materialises, a milestone turns Red, or a blocker has been unresolved beyond the configured SLA
- Produces structured escalation briefs for programme risks that require human decision: the issue description, its impact on the programme timeline and on dependent programmes, the decision required (scope change, resource addition, timeline extension, or issue acceptance), the options available with their trade-offs, and the recommended escalation path — which human should make the call and by when.
- Manages the escalation follow-through: tracks whether the escalated issue received a decision within the required timeframe, re-escalates to the next level if the decision owner has not responded, and updates the risk log with the decision outcome and any residual risks.
Output: Escalation brief — issue, impact, decision required, options with trade-offs, recommended path, decision owner, and response deadline. Follow-up re-escalation if no response within SLA.
Change Rollout Management
For all operational and system change deploymentsInvoked when: a change is approaching its go-live date and requires pre-rollout risk assessment and readiness confirmation
- Produces a change readiness assessment 5 days before any planned go-live: are all pre-conditions met (testing complete, training delivered, rollback plan documented, regulatory notifications made where required, impacted teams confirmed ready)? Any pre-condition not met becomes a go-live risk flagged to the programme sponsor for a go/no-go decision.
- Sequences change rollouts to avoid conflicts — two system changes to the same platform in the same week, or a go-live scheduled during a high-volume operational period (month-end, festival season) where a rollback would be more disruptive. Flags scheduling conflicts to the COO AI for rescheduling decision.
- Post-rollout: monitors the first 72 hours of a change for stabilisation signals (error rates, processing volumes, escalation spike) and produces a go-live health report. If stabilisation signals indicate a problem, escalates to the relevant teams and the COO AI immediately rather than waiting for the scheduled post-implementation review.
Output: Pre-rollout readiness assessment, scheduling conflict flags, go/no-go recommendation for sponsor, and 72-hour post-rollout health report.
Hard guardrails
Will notApprove scope changes, budget increases, or resource additions. All changes to programme scope or resource require the programme sponsor's sign-off, regardless of how small the change appears operationally.
Will notCommit to delivery timelines on behalf of any functional team. The agent tracks and reports progress; commitments are made by the functional heads who own the delivery capacity.
Will notAuthorise a change go-live. The go/no-go decision is the programme sponsor's or COO AI's call. The agent provides the readiness assessment; the human makes the call.
Known limitations
Programme tracking quality depends entirely on task status being updated in the project management system by the owning teams. If teams do not update their task status promptly, the agent's RAG statuses will be based on stale data and the escalation will be late relative to the actual problem.Build task status update compliance into team performance metrics — the percentage of tasks with up-to-date status should be a KPI reviewed in the weekly ops review. Stale task data is a programme governance failure, not a PM agent limitation.
The agent cannot resolve conflicts between competing functional priorities — it can identify the conflict and present the options, but resolving it requires a human with authority over both teams. Where the same conflict recurs across multiple programmes, the root cause is usually a structural capacity or prioritisation gap that requires a strategic rather than a per-programme fix.Track recurring conflict patterns in the escalation log — the same team appearing as a blocker across three or more programmes in a quarter signals a structural capacity or prioritisation issue that the COO AI should address at the strategic planning level.
Important Reads
Learn more about how to deploy Ops Programme Manager AI to your lending workflow.
- Use case #0001Cross-team coordination: how Ops PM AI keeps origination, credit and disbursement in syncA home loan application involves six functions, fourteen handoff points, and thirty-two SLA obligations before a rupee disburses. When origination slips on a document checklist, credit waits. When credit extends a query, legal sits idle. When legal misses a title search, disbursement cannot proceed. The Ops Programme Manager AI tracks every dependency across every function in real time — not by attending status meetings, but by monitoring the handoff events that determine whether the operation runs on time.Read article →
- Use case #0002Change rollout management: how Ops PM AI deploys new RBI guidelines across processesWhen the RBI issues a new circular affecting collections conduct, the change does not just need to be acknowledged — it needs to be deployed. Scripts updated, agent training completed, IVR flows modified, WhatsApp templates approved, compliance logs configured. Each of these is a task with an owner, a dependency, and a deadline derived from the circular's effective date. The Ops Programme Manager AI converts the regulatory change into a project plan and tracks every task to completion before the effective date.Read article →
- Use case #0003Escalation management: the decision tree Ops PM AI uses to route exceptionsAn exception that reaches the wrong desk is an exception that takes three days to resolve. An exception that reaches the right desk — with the right context, at the right level of authority, with the resolution options already outlined — takes three hours. The Ops Programme Manager AI routes every exception in the lending operations pipeline through a decision tree that determines: what kind of exception this is, who has the authority to resolve it, what information they need, and how long they have before the delay cascades into a pipeline failure.Read article →