AI / ML Engineer focused on Generative AI — RAG, tool-calling agents, and shipping LLM applications end-to-end. Strong software-engineering foundation (Python, FastAPI, C++) with a background spanning XR/robotics, Android, and computational geometry.
🌐 Website: https://smh997.github.io/ · 💼 LinkedIn: https://www.linkedin.com/in/mohammad-hashemi1997/
Working through a hands-on, RAG-first path into AI engineering and shipping a portfolio of production-style LLM apps.
| Project | What it demonstrates |
|---|---|
| 🔎 RAG Document Copilot | Retrieval-augmented generation over PDFs with grounded, page-cited answers. FastAPI · Cohere embeddings · Chroma vector store · pypdf. Refuses to answer outside its sources — real RAG, not a chatbot. |
| ⚽ FPL Tool-Calling Agent | Natural-language agent that plans and calls real API tools in a loop over the live Fantasy Premier League API. Cohere tool-calling · multi-step reasoning · FastAPI. |
Next up: containerized deployment of a mini AI product, and continued work on agent patterns.
GenAI stack: RAG · embeddings & vector search · tool-calling / agents · prompt engineering · Cohere · Chroma
- 🥽 XR / AR-VR — still building immersive experiences; open to collaboration and open-source here. 👉 XR Learning Journey
- 📱 Android — Kotlin, Jetpack Compose (shipped on Google Play)
- 📐 Computational geometry & competitive programming — 5+ years, algorithm tutoring/TA experience
| 🧩 Category | 📦 Sample Projects |
|---|---|
| 🤖 AI / GenAI | 🔎 RAG Copilot, ⚽ FPL Agent |
| 🥽 AR/VR | 🎲 Backgammon VR, 🏠 VR Room, 🚩 WayPoint Creator |
| 📱 Android | 🧭 Amsterdam, 📚 BookExplorer, 🔤 Unscramble, 🧾 Inventory, 💧 Water Me, 🪐 Mars Photos, 📘 30 Days of Vocabulary |
| 📊 Classical ML | 💳 Transactions Prediction, 🎧 Audiobook Segmentation, 🌐 Community Detection |
| 🧠 Problem Solving | 💡 Competitive Programming |
| 🖧 Backend | 🌱 Bilche |
