AI Audit Reports
A lightweight, AI-powered analytics MVP built in one week on top of Etrends' Laser Audit Reporting System. This experimental tool connects to the existing SQL views to generate instant, human-friendly summaries, custom reports, and conversational insights—helping audit teams discover patterns and answers that the original dashboards can't. Designed purely as a minimum viable product to test AI-driven reporting, not as a full-featured release.
Background
Etrends' Laser Audit Reporting System (LARS) is an established internal audit platform used by organizations to centralize audit planning, scheduling, work papers, findings, and live dashboards. While powerful, teams needed more flexibility: custom summaries, ad-hoc reports, and the ability to ask questions of their data without SQL or report builders.
What We Built
An AI-powered analytics layer built in one week as an MVP. It connects directly to existing LARS SQL views and adds conversational analysis, instant summaries, and custom report generation without touching the core product.
Summary Generator Feed it any SQL view and get clean, readable summaries in seconds. "Compliance looks like this across all locations this quarter." "Site X lagged in observation closure." Copy-paste ready for slides and emails.
Custom Report Generator Describe the analysis you need in plain English: department-wise risk breakdowns, location comparisons, trend analysis. The system queries the data, generates charts, and writes the summary. No manual exports or spreadsheet work.
Conversation Chat Ask your audit data direct questions: "Which locations have the most overdue actions?" "Summarize last month's top risks across SBUs." Get context-aware, natural responses. Follow up and drill deeper without writing a single query.
Why an MVP?
- Built in one week to prove what's possible before committing to full integration
- Plugs into existing SQL views, no re-architecture of LARS required
- Designed to spark feedback from real audit teams before building the full product
Results
- 97% time reduction: reports that took 6 hours now take 8 minutes
- 10x output capacity: audit teams can generate vastly more analysis with the same headcount
- Validated demand for AI-driven reporting before investing in a full product build
Tech Stack
Next.js, NestJS, SQL views, OpenAI, LangChain, LangGraph, Tailwind CSS