I build AI agents, LLM-powered tools, and voice systems that actually ship —
backed by open-source code with 5,700+ downloads.
📍 Islamabad, Pakistan · 📧 yasirabdali6@gmail.com
| Area | Tools & Frameworks |
|---|---|
| LLM Agents & Orchestration | LangGraph, LangChain, LlamaIndex, FastMCP |
| Voice & Multimodal AI | ElevenLabs, Fish-Speech, ChatterBox, Gemini |
| Model Fine-tuning | LoRA, QLoRA, Unsloth, PEFT, Transformers |
| Deployment & Infra | RunPod serverless, Docker, FastAPI, AWS |
| RAG & Memory | Chroma, semantic chunking, hybrid retrieval |
🔌 NotebookLLM MCP Server · PyPI · 5.7K+ Downloads · Listed on LobeHub
Production-ready MCP server for efficient LLM interactions with Jupyter Notebooks. Converts .ipynb files to plain text before sending to the LLM — cutting token usage dramatically.
pip install notebookllm_mcpLangGraph agent with cloned voice (ElevenLabs), image understanding (Gemini), and RAG memory (Chroma). Accepts text, voice, and image input — responds in audio or text with a persona-aligned voice.
Stack: LangGraph · Telegram API · Gemini · ElevenLabs · Chroma
LangGraph + Twilio agent handling text, voice, and image inputs. Outputs contextual, persona-grounded responses using vector memory.
Stack: LangGraph · Twilio · Gemini · Chroma
Fine-tuned NuminaMath-7B with LoRA to generate executable Python code from LaTeX math. Includes synthetic data generation pipeline and SymPy-based validation.
Stack: Unsloth · QLoRA · SymPy · HuggingFace
Deployed state-of-the-art open-source TTS models as serverless endpoints on RunPod using Docker multi-stage builds with chunked long-text processing.
Autonomous research agent with real-time voice/video assistance, browser automation, full-stack software generation, and MCP server extensibility.
| Metric | Value |
|---|---|
| PyPI Package Downloads | 5,700+ |
| Published PyPI Packages | 2 (notebookllm, notebookllm_mcp) |
| GitHub Repositories | 51 |
| Remote Work Experience | 18 months (Teamlift, US) |
| Kaggle Status | Expert |
| NER pipeline scale | 100,000+ job descriptions |
Junior Data Scientist — Teamlift (US Remote · Jan 2024 – Dec 2024)
Built end-to-end automated data pipelines with Human-in-the-Loop design. Engineered prompts for skill extraction and implemented function calling for ambiguous skill verification.
ML Intern — Teamlift (US Remote · Jul 2023 – Dec 2023)
Web scraping at scale (Scrapy, Playwright, BeautifulSoup), NER on 100K+ job descriptions, time-series forecasting for job trend prediction.
- B.Sc. Software Engineering — NUML, Islamabad (Graduating 2026)
- Advanced LLM Bootcamp — NUST & Sky Electric (2024)
- Google: Generative AI Fundamentals · Responsible AI · LLMs
- DataCamp: Data Scientist Track · Kaggle Expert
I'm available for freelance projects in:
- Custom LLM agent systems (LangGraph, LangChain)
- Voice AI & multimodal agents
- LLM fine-tuning (LoRA/QLoRA/Unsloth)
- MCP server / tool development
- RAG chatbots & document QA systems
📬 Email: yasirabdali6@gmail.com
💼 Upwork: upwork.com/freelancers/~0136d3079a657bfcbb
💼 Fiverr: fiverr.com/sellers/yasirrazaa
🔗 LinkedIn: linkedin.com/in/yasir-raza
🔗 Kaggle: kaggle.com/yasirabdaali
🔗 PyPi: pypi.org/user/yasirabdali6