We are looking for an experienced Python Backend Engineer to design and implement a RAG (Retrieval-Augmented Generation) based AI bot that deeply understands our databases, procedures, and application flows, and can assist with all types of database-related tasks. The goal is to build an internal intelligence bot that can: Answer natural-language questions about our databases Explain stored procedures, jobs, and workflows Generate architecture diagrams, ER diagrams, and flow charts Help write and validate SQL queries Assist in documentation, reporting, and impact analysis This bot will be used by developers, DevOps, and analysts as a single interface to understand and interact with our systems. Core Objectives 1. Knowledge Ingestion (RAG Layer) The bot should ingest and index: Database schemas (tables, columns, relationships) Stored procedures, functions, triggers, jobs Existing SQL scripts Application documentation Admin panel procedures and workflows Sample reports and queries This knowledge should be searchable and retrievable using vector embeddings. 2. Database Intelligence The bot must be able to: Explain table relationships and dependencies Answer questions like: “Which tables are impacted if I change column X?” “How does this stored procedure work step-by-step?” “Which procedures write to this table?” Generate optimized SQL queries from natural language Validate queries against schema rules Help analyze performance and indexing 3. Architecture & Flow Visualization The bot should generate or assist in generating: ER Diagrams Database schema diagrams Procedure execution flows Data movement flow charts High-level system architecture diagrams 4. Dev & Ops Support The bot should assist with: Data extraction for reports Root-cause analysis using DB logs and queries Explaining production vs staging differences Helping new developers onboard faster Answering questions about deployment and data flows Expected Bot Capabilities (Examples) ✔ “Explain how the daily_summary procedure works” ✔ “Generate an ER diagram for the finance schema” ✔ “What happens if we delete data from this table?” ✔ “Write a query to get monthly exchange closing data” ✔ “Show data flow from ingestion to reporting” ✔ “Create a flow chart of this admin procedure” Tech Stack (Preferred, Flexible) LLM: Gemini RAG Framework: LangChain / LlamaIndex Vector DB: Pinecone / Weaviate / FAISS / Chroma Databases: MySQL Backend: Python (FastAPI preferred) Auth: Azure AD / IAM / RBAC Deployment: Azure (VM) Diagram Generation: Mermaid / PlantUML Security & Access Read-only DB access for ingestion No data used for model training Private embeddings and enterprise-grade security On-prem / private cloud deployment only Deliverables Working RAG-based AI bot (API or UI) Ingestion pipelines for DB + docs Query generation and explanation module Diagram and flow chart generation Documentation + onboarding guide Chat Interface