← Agent catalogue

AI Agent Profile · LendingIQ · Agent #87 · MIS

MIS & Reporting Agent AI

Function: MIS ExecutiveInvoked via: daily 6 AM data pull · on-demand report request · SLA breach alertRuntime: AWS Bedrock · ap-south-1Model: Claude Sonnet 4Context window: 200K tokens

DivisionLending Operations

Resume

What this agent does

The MIS & Reporting Agent AI replaces the MIS executive function entirely — pulling data from every operational system at 6 AM, assembling the daily ops pack that lands in the COO's inbox by 7 AM, maintaining a live SLA dashboard that refreshes every 15 minutes throughout the day, and producing the weekly performance report that ops leadership uses to run their Monday review. It covers origination TAT, disbursement volumes, collections efficiency, QC error rates, and branch-level performance across every active location — giving leadership a single, consistent view of operational health without a human intermediary in the reporting chain.

Primary functions

Daily Ops Pack

Daily · data pulled at 6 AM · delivered to COO by 7 AM

Invoked when: the daily 6 AM scheduled run begins — all source systems queried and the ops pack assembled and dispatched before 7 AM

  • Queries all source systems at 6 AM and assembles the daily ops pack in a standardised format — a single document that gives the COO a complete view of the prior day's operations before their first meeting. The pack covers: (1) origination — applications received, approved, and declined in the prior day, approval rate, and the top 3 decline reasons; (2) disbursement — loans disbursed (count and value), average disbursement TAT for the day vs the 7-day average, and any disbursements that breached the TAT SLA; (3) collections — DPD bucket movements (borrowers entering and exiting each bucket), same-day collection efficiency (payments received as a percentage of payments due), and recovery on written-off accounts; (4) quality — QC Agent AI's error rate for the prior day's sample; (5) operations — active applications in each stage, count of applications on hold and hold reason summary, and SLA breach count for the prior day.
  • Flags anomalies automatically in the pack — any KPI that has moved more than 15% from the prior week's equivalent value receives a bold anomaly note immediately below the metric: "Disbursement TAT: 38 hours (↑ 22% vs last Tuesday's 31 hours — review stage 4 processing). " The anomaly note includes the specific metric, the magnitude of the movement, the direction, and the stage or category most likely responsible, to save the COO the step of identifying where in the operations the movement originated.
  • Calibrates the narrative to the COO's reading pattern — the ops pack opens with a 3-sentence executive summary (yesterday in one sentence, the key anomaly in one sentence, and the one action recommended for today in one sentence), followed by the full metric tables. The COO who reads only the executive summary has the essential context for their morning decisions; the COO who reads the full pack has the complete operational picture.
Output: Daily ops pack — 5-section operational summary with anomaly flags and 3-sentence executive summary. Delivered to COO and ops head by 7 AM every business day. Prior day's data as of 6 AM data pull; same-day data visible in the live dashboard.

TAT Reports & Live SLA Dashboard

Real-time dashboard · refreshed every 15 minutes · branch-level TAT reports weekly

Invoked continuously for the live dashboard; weekly TAT report for branch performance runs every Monday alongside the weekly performance report

  • Maintains a live SLA dashboard that refreshes every 15 minutes throughout business hours — showing all active applications with their current stage, elapsed time at stage, cumulative TAT, and SLA status (green / amber / red based on the SLA thresholds from the Onboarding SLA Agent AI). The dashboard is the ops head's intraday tool — the daily ops pack tells them what happened yesterday; the live dashboard tells them what is happening now. Branch managers access a filtered view of their own branch's live status.
  • Tracks TAT across the full origination-to-disbursement journey for every active application — broken down into the individual stage times (KYC, QC, credit assessment, sanction, documentation, disbursement) so that the ops head can see where time is accumulating rather than just the total TAT. A total TAT of 72 hours that breaks down as 48 hours in credit assessment and 24 hours across all other stages is a different problem from a 72-hour TAT that is distributed evenly — the breakdown determines the intervention.
  • Produces the weekly branch TAT report covering all 50 branches — ranking branches by their average origination-to-disbursement TAT for the week, identifying the fastest and slowest branches, and flagging branches whose TAT has deteriorated more than 20% relative to the prior 4-week average. The branch comparison surfaces whether TAT problems are systemic (affecting all branches — likely a platform or process issue) or local (affecting specific branches — likely a capacity or supervisory issue).
Output: Live SLA dashboard — all active applications with current stage, elapsed time, and SLA status. Refreshed every 15 minutes. Weekly branch TAT report — all branches ranked by average TAT, with anomaly flags for branches deteriorating more than 20% vs prior 4-week average.

Weekly Performance Report & On-Demand Reports

Weekly (Monday 8 AM) + on-demand within 10 minutes

Invoked when: Monday weekly cycle runs, or an on-demand report request is received from an authorised user

  • Assembles the weekly performance report every Monday by 8 AM — a comprehensive view of the prior week's operations across all functions, formatted for the ops leadership team's Monday review meeting. The weekly report covers the same 5 operational areas as the daily pack but with a week-on-week comparison, a 4-week trend for each KPI, and a commentary section that contextualises the week's movements in relation to the prior period. The weekly report includes the branch performance ranking, the agent productivity summary (applications processed per agent per day by branch), and the quality summary (QC error rate trend and the top 3 error categories for the week).
  • Delivers on-demand reports within 10 minutes of a request from an authorised user — COO, ops head, branch head, or any leadership team member with MIS access. On-demand reports can be standard templates (the daily pack for a prior date, a specific branch's 30-day performance, a specific product's TAT breakdown) or ad-hoc queries (all applications in credit assessment for more than 5 days, disbursement volumes by borrower segment for the prior quarter). Ad-hoc queries that require data not available in the standard MIS are flagged as requiring a custom data pull from the technology team.
  • Maintains a report library — the archive of all delivered reports, available for retrieval by date, report type, and recipient. The report library is the evidence base for audits and management reviews where historical operational performance data is needed; every report delivered is retained in its original format with the timestamp of delivery.
Output: Weekly performance report — 5-function operational summary with 4-week trend and branch comparison. Delivered Monday 8 AM. On-demand reports within 10 minutes for standard templates; ad-hoc data pull requests flagged to technology for custom extraction. Report library maintained for audit and management review access.

Knowledge base

LOS — Application, Disbursement, and Stage Event Data

The primary operational data source — application volumes, stage TATs, disbursement counts and values, decline rates and reasons. Queried at 6 AM for the daily pack and in real-time for the live dashboard.

Collections System — DPD and Recovery Data

DPD bucket data, collection efficiency metrics, and written-off account recovery — included in the daily ops pack and weekly performance report.

Branch Management System — Branch-Level Data

Branch-level volumes, TATs, and agent productivity — the source for the weekly branch comparison pack and the 50-branch TAT ranking.

Report Templates and Format Standards

The standardised formats for all recurring reports — daily ops pack, weekly performance report, branch TAT report, and on-demand standard templates. Maintained by the ops head and updated when reporting requirements change.

Historical MIS Report Archive

All prior delivered reports — retained for audit, management review, and on-demand retrieval of historical performance data.

Pre-Training — MIS and Operations Reporting Knowledge

Management information system design, TAT and SLA reporting methodology, and operational KPI frameworks for lending operations up to knowledge cutoff.

Hard guardrails

Will notMake any operational decision or recommendation. The MIS pack surfaces what the data shows — what to do about it is the COO's and ops head's decision. Where an anomaly is flagged, the note identifies the likely source; it does not prescribe a response.
Will notCorrect source system data discrepancies. Where the LOS and another system show conflicting data for the same metric, both figures are reported with a discrepancy flag, and the technology team is notified. The agent does not choose between conflicting data points or make assumptions about which is correct.
Will notDeliver reports to unauthorised recipients. The recipient list for each report type is maintained by the ops head; the agent delivers to the specified list and does not add recipients or forward reports without explicit authorisation.
Will notDelay the 7 AM daily pack delivery. If a source system is unavailable at the 6 AM pull, the pack is assembled from the available systems and delivered at 7 AM with a note identifying which system's data is missing. An incomplete pack delivered on time is preferable to a complete pack delivered late for the COO's morning schedule.

Known limitations

The 6 AM data pull captures data as it exists in the source systems at that moment — it does not capture transactions processed between 6 AM and the start of business. For a high-volume lending operation, the gap between the 6 AM snapshot and the 9 AM business start may include a meaningful number of overnight system processes. The live dashboard addresses this gap for intraday decisions; the daily ops pack is a prior-day summary, not a real-time balance.Clearly label the daily ops pack as a prior-day summary based on a 6 AM data cut — preventing the COO from treating the pack as real-time data for intraday decisions. Real-time data should be read from the live dashboard, not from the ops pack.
The anomaly threshold (15% movement triggers a flag) is a fixed parameter — it may be too sensitive for metrics with high natural volatility (daily application volumes vary significantly by day of week) and too insensitive for metrics with low natural volatility (disbursement TAT should be stable). A single threshold applied to all metrics produces false positives on volatile metrics and may miss meaningful movements in stable ones.Calibrate the anomaly threshold per metric — using the metric's historical volatility (coefficient of variation) to set a threshold that flags genuine anomalies rather than normal fluctuations. A disbursement TAT flag threshold of 10% is tighter than appropriate for application volume but appropriate for a metric that should be stable.
Agent Profile · MIS & Reporting Agent AI · LendingIQ · Agent #87Last updated April 2026 · For internal use

Important Reads

Learn more about how to deploy MIS & Reporting Agent AI to your lending workflow.