From an idea to a useful AI product — thoughtfully built.
I'm an Applied AI engineer who enjoys building useful, tangible products with AI.
I began with Electrical Engineering, microcontrollers, embedded systems, and robotics. I then moved into Computer Vision and Deep Learning, before focusing on AI software. That journey still shapes how I work: I care about connecting technology to real problems and creating things that people can actually use.
- 🤖 LLM applications: RAG, agentic workflows, prompt engineering, and AI assistants
- 👁️ Computer Vision: detection, segmentation, classification, OCR, and edge deployment
- ⚙️ AI backends: FastAPI services, model serving, APIs, and data pipelines
- 🐳 Deployment & infrastructure: Docker, PostgreSQL, MinIO, Linux, and GPU/edge deployment
- 🔄 Workflow automation: Selenium, n8n, RPA, document classification, and business-process automation
- 💡 Product prototyping: turning ideas into interactive, testable experiences
|
An interactive pixel-art world where Persian poets answer questions about their poems and ideas using Agentic RAG.
|
✏️ Pencil AIAn iPad-friendly AI drawing experience where users draw a character, the system understands it, and the character becomes interactive.
|
🏙️ CodeCityExplore any codebase as an interactive 3D city — classes as buildings, folders as districts. Select a building to highlight dependencies and preview source, entirely offline without executing project code.
|
A personal AI round table — run multiple local or OpenAI-compatible models together with independent or deliberative conversations, editable roles, MCP tools, and usage tracking.
|
Before focusing on LLM applications, I worked extensively on Computer Vision and Deep Learning: object detection, segmentation, classification, model optimization, and deployment on servers and edge devices.
Some of my earlier open-source work includes OpenCV ROI Selection, Awesome CARLA, and several Computer Vision and Deep Learning resources.
- 📏 Reliable evaluation methods for RAG and LLM applications
- 🧭 Agentic systems that can solve real workflows
- 🧩 Better multimodal interaction between people and AI
- 🎨 AI products with simple, tangible, and enjoyable user experiences
AI: LLMs · RAG · Agents · Embeddings · Computer Vision · OCR · Multimodal AI
Backend & Deployment: FastAPI · Docker · PostgreSQL · MinIO · Linux · Ollama · vLLM · LiteLLM · llama.cpp
Automation: Selenium · n8n · RPA · Document Processing
