Moiz Ahmed Mansoori
AI Assistants, RAG Chatbots & Business Workflow Automation
New to Contra
Moiz Ahmed is ready for their next project!
QueryMind lets you query a real database using plain English.
Type a question, get an answer — no SQL knowledge needed.
Built a production-grade AI agent using LangGraph that converts
natural language to SQL, executes it on a real PostgreSQL database
with 100,000+ rows, and automatically corrects errors in a retry
loop if the query fails.
Key features:
7-node LangGraph agent with conditional routing and self-correction
pgvector semantic search for intelligent schema retrieval
Full observability dashboard with real-time metrics and query history
Trace replay — see every step the agent took to answer your question
Deployed live: Next.js frontend on Vercel, FastAPI backend on Render
Tech: LangGraph · Groq LLaMA 3.3 70B · PostgreSQL · pgvector ·
FastAPI · Next.js · Tailwind CSS
Github: https://github.com/moiz-mansoori/QueryMind-NL2SQL-Agent-System
0
30
Built AIRO, a 6-agent ML automation system using LangGraph and Groq LLaMA 3.3. The system automates the entire ML pipeline from data ingestion to final report generation. Ran 18 experiments achieving RandomForest f1_macro of 0.7329. Features parallel training, MLflow tracking, SHAP explainability, and a dark Streamlit UI with live log streaming.
2
137
Built a deployed AI tool that lets users upload research papers and query them using natural language. RAG pipeline with vector search retrieves precise context-aware answers. Features PDF parsing, AI summaries, topic modeling, and citation extraction. Live on Render.
6
3
230
Built a full-stack Todo app across 5 phases — CLI to cloud-native. Phase 3 adds AI chatbot via LangChain and Groq for natural language task management. Phase 4 containerized with Docker and Kubernetes. Phase 5 deployed on Vercel with Neon PostgreSQL. 4 live deployments.
1
140
Built a deep learning system to detect brain tumors from MRI scans using VGG16 transfer learning. Integrated Grad-CAM heatmaps for explainable AI output. Achieved 94.9% confidence with 4-class classification. Full pipeline from raw MRI input to visual diagnosis output.
1
133
Built an empathetic mental wellness chatbot using LLaMA 3.3 and LangChain. Features mood-based conversations, real-time sentiment analysis, ChromaDB vector memory, breathing exercises, and journaling prompts. Deployed live on Streamlit Cloud.
1
105