Looking for a job

Muhammad Ehtasham

@muhammadehtasham

Message

I build production AI applications with React, FastAPI, and custom models.

Pakistan

Message

I've built production AI products for Instabill.co, SimbaRise.com, and PinballRace.com, combining React 19 frontends, FastAPI backends, and AI integrations.

For Instabill.co, I developed a GCP Gemini backend that extracts and validates bank-statement and invoice data for high-volume financial documents. At SimbaRise.com, I built subscription access controls and API integrations for premium ChatGPT services.

I also engineered PinballRace.com as a high-frequency gaming platform and integrated YOLOv12 for automated game-state verification. My freelance work includes ETL pipelines, OpenAI-powered JSON/XML generation, private local-LLM document chat systems, and domain-specific LLM fine-tuning.

In open-source medical imaging research, I modified HoVer-Net on the PanNuke dataset to detect hyperactive brain tumors with 89% accuracy. I also mentor junior AI students at university hackathons and am preparing research on ultralightweight transformer LLMs.

Experience

Work history, roles, and key accomplishments

Current

Full Stack AI Engineer

Current

Freelance

Aug 2024 - Present (2 years 1 month)

Developed financial AI backend for Instabill.co using GCP Gemini, built AI SaaS platform for SimbaRise.com, and engineered gaming platform PinballRace.com with React 19 and FastAPI.

GCP Gemini React 19 fastAPI YOLOv12 API Integration Financial Document Processing

AI Researcher (Medical Imaging)

Open Source

Jan 2025 - May 2025 (4 months)

Modified HoVer-Net architecture on PanNuke dataset achieving 89% accuracy in detecting hyperactive brain tumors and automated quantitative analysis for H&E and IHC modalities.

HoverNet PanNuke Medical Imaging Deep Learning Segmentation

Custom AI Solutions & Pipeline Engineer

Freelance

Dec 2023 - Aug 2024 (8 months)

Designed ETL pipelines for e-commerce metadata, built privacy-first chat systems with local LLMs, and fine-tuned custom LLMs on proprietary datasets.

ETL OpenAI API Ollama GPT4All LLM Fine tuning Data Engineering

AI Architecture Debugging Specialist

Freelance

Apr 2023 - Dec 2023 (8 months)

Specialized in rapid debugging of broken AI training pipelines, fixing PyTorch/TensorFlow errors and optimizing training loops with under 24-hour turnaround.

PyTorch TensorFlow Debugging Training Optimization