RCA agent + Insights agents for Dashboards

I’m looking for someone who has **already built AI agents for business dashboards** — particularly applications involving **sales, operations, inventory, supply chain, or similar business data**. This is **not a generic chatbot project**. I’m specifically looking for someone who has experience building agents that can look at structured business data, identify meaningful changes/anomalies, investigate the underlying drivers, and generate actionable insights. ### What I want to build An AI intelligence layer on top of a business dashboard with agents such as: **1. Insights Agent** * Identifies important changes/anomalies in business metrics * Detects trends, deviations, outliers and unusual patterns * Prioritizes the most important issues * Explains the impact quantitatively * Works across dimensions such as SKU, product, region, channel, customer, etc. **2. RCA Agent** * Investigates *why* an identified issue happened * Decomposes metrics into underlying drivers * Traces causes across multiple dimensions * Quantifies the contribution of different drivers * Provides evidence and confidence for the RCA * Produces an explainable causal/drill-down tree For example: - Inventory increased ₹4 Cr → 43% due to demand deterioration → concentrated in South India → primarily 5 SKUs → demand fell after promotion ended → forecast was not adjusted. The goal is to make this **interactive and explainable**, rather than simply having an LLM generate commentary from charts. ### Who I'm looking for You should have **hands-on experience building something similar**, preferably: * AI agents over structured business/operational data * RCA / root-cause analysis systems * AI insights or anomaly detection for dashboards * Sales analytics / operations analytics / inventory / supply-chain applications * Agentic workflows using Python + LLMs * SQL / data analysis / APIs * Experience grounding LLM outputs in actual data rather than allowing hallucinated explanations **Prior experience is much more important than theoretical knowledge.** Please share **1–2 examples of actual applications you have built**, ideally with screenshots, demo links, GitHub, or a brief explanation of the architecture. ### Engagement Initially I want to build a **working prototype**, with potential for longer-term development if the fit is good. I'm open to either an hourly engagement or a fixed-price prototype. **When applying, please answer:** 1. Have you built an AI insights/RCA agent for a dashboard before? 2. What was the application/domain? 3. What exactly did the agent do? 4. What was the underlying data and architecture? 5. Please share a demo, screenshots, GitHub, or other evidence of the work if available. Please **do not apply with a generic AI-agent portfolio**. I'm specifically looking for someone who has built **business intelligence/analytics agents over structured data**.