I ship production AI systems — multi-tenant RAG platforms, multi-agent pipelines, and PyTorch computer vision models. Two client systems solo-built from architecture to deployment, both live today. Recent BS in Data Science, FAST-NUCES Lahore (2026) · Paper published to SSRN.
📍 Lahore, Pakistan · Open to remote & on-site roles
|
LLM / RAG Systems Multi-tenant chatbot SaaS, per-tenant vector-store isolation, multi-agent content pipelines |
Computer Vision + XAI PyTorch late-fusion architectures, 10-method explainability suites, dash-cam crash detection |
|
Production Backends FastAPI / Flask at scale, multi-service Nginx deployments, Docker on AWS |
Data Pipelines PySpark on AWS EMR at 18M+ record scale, ML analytics, automated ingestion |
Languages
Python · TypeScript · JavaScript
ML / AI
RAG · LangChain · Multi-Agent · OpenAI API · Gemini API · ChromaDB · FAISS · Pinecone · Sentence-Transformers
Backend / Frontend
FastAPI · Flask · React · Next.js · PostgreSQL
Cloud & DevOps
AWS · Docker · Nginx · Linux · Git
Client Work @ Wesolvd (code available on request)
| Project | What it is | Stack |
|---|---|---|
| LeadsRep ChatBot | Multi-tenant RAG SaaS live at leadsrep.co — hand-rolled RAG pipeline, per-tenant ChromaDB isolation, embeddable widget, 90-route FastAPI backend | FastAPI · ChromaDB · Next.js · Docker |
| KataLog (Inventory Planner) | Multi-store Shopify inventory platform — replenishment forecasting, cross-store mirroring microservice · warehousekatalog.com | FastAPI · React 19 · Shopify API |
| AI Content Refinement | Multi-agent pipeline (Engineer / Creator / Evaluator) with RAG over Pinecone, orchestrated via n8n | LangChain · Pinecone · n8n |
───────────────── [ EOF ] ─────────────────