shahh19.onlinePortfolio
Hassam Shah AI Engineer
Engineering intelligent systems that learn, reason, and scale.
I design, train and deploy machine learning systems — spanning deep learning, computer vision, natural language processing, intelligent automation and AI-powered applications.
- AI
- ML
- CV
- NLP
- LLMs
01About
From data to deployed intelligence.
I'm Hassam Shah, an AI engineer who enjoys the full arc of building intelligent systems — framing the problem, shaping the data, training models and getting them to work reliably in the real world.
My interests sit where research meets engineering: computer vision, language models, intelligent automation and the infrastructure that turns a promising notebook into a dependable product.
I care about clear evaluation, simple architectures that are easy to reason about, and software that people can actually use.
{ "name": "Hassam Shah", "role": "AI Engineer", "focus": ["ML", "Deep Learning", "CV", "NLP"], "builds": ["models", "pipelines", "AI apps"], "location": "add in js/data/site.js", "open_to_collab": true }
Designing AI systems
Architectures that fit the problem, not the hype.
Training models
From classical ML to deep neural networks.
Computer vision
Systems that understand images and video.
Deploying ML
APIs, containers and reliable inference.
>_How I work · ML lifecycle
Frame
Define the problem, constraints and what success means.
Data
Collect, clean and understand the data first.
Model
Start simple, then iterate on architecture and training.
Evaluate
Measure honestly against the right metrics.
Deploy
Ship behind an API, monitor and improve.
02Skills
A stack built around the model lifecycle.
Tools grouped by what they do in a real system — from research and training to data and production.
03Experience
Where the work happened.
04Projects
Selected work.
Systems across vision, language, prediction and automation. Open any project for the full case study.
05AI Lab
What I'm exploring.
Open research threads and experiments — the topics I'm reading about, prototyping and testing.
06Neural Playground
Watch a network learn, live.
A real neural network trained in your browser with backpropagation — not an animation. Pick a dataset, hit train, and hover the input space to run inference on any point.
- Epoch
- 0
- Loss · BCE
- —
- Accuracy
- —
- Params · state
- — · Ready
07Engineering Activity
Open source & repositories.
Live public data from GitHub: repositories, languages and recent activity.