Skip to content
View ha7n23's full-sized avatar

Block or report ha7n23

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ha7n23/README.md

Hi, I'm Huzaifa 👋

I'm a Computer Science graduate from the University of Southampton, focused on AI Engineering, GenAI applications, RAG systems, and production-style LLM applications.

My current portfolio focuses on building practical LLM systems beyond notebooks, including API design, retrieval pipelines, tool-using agents, evaluation, Docker, CI/CD, and safety controls.

AI Engineering Focus

  • Retrieval-Augmented Generation RAG systems
  • LLM application development with FastAPI
  • Vector databases and semantic search
  • Tool-using agents and controlled workflows
  • LLM evaluation, citation validation, and quality gates
  • Docker, GitHub Actions, and production-style project structure

Featured Projects

Banking Knowledge RAG Assistant

Production-style banking RAG assistant for customer-support style questions.

Built with FastAPI, Chroma, sentence-transformer embeddings, hybrid retrieval, metadata filtering, query rewriting, reranking, grounded LLM generation, source citations, answer evaluation, Docker, GitHub Actions CI, and a lightweight browser UI.

Key skills: RAG, FastAPI, Chroma, vector databases, LLM evaluation, Docker, CI/CD

View Repository


Banking Support Agent

Controlled banking support agent demonstrating safe tool-using AI workflows.

The agent uses intent routing, typed mock banking tools, grounded LLM responses, and confirmation-gated actions before sensitive workflows such as dispute creation.

Key skills: AI agents, tool calling, FastAPI, workflow control, safety, Docker, testing

View Repository


Multi-Provider LLM Starter Kit

Reusable LLM starter kit supporting Gemini, Groq, and Mistral through a clean provider abstraction layer.

Includes CLI usage, FastAPI endpoints, streaming responses, structured JSON outputs, environment-based configuration, tests, Docker, GitHub Actions CI, and documentation.

Key skills: LLM APIs, provider abstraction, streaming, structured outputs, FastAPI, Docker

View Repository

Technical Stack

Languages: Python, SQL
AI/LLM: RAG, LLM APIs, prompt engineering, embeddings, vector search, agents
Frameworks: FastAPI, Pydantic, Chroma
Engineering: Docker, GitHub Actions, pytest, CI/CD, structured logging, environment-based configuration
Data/Analytics: pandas, Power BI, data analysis, reporting

Current Direction

I'm currently preparing for AI Engineer, GenAI Engineer, LLM Application Engineer, and Applied AI Engineer roles, with a focus on building reliable, evaluated, and production-style LLM applications.

Pinned Loading

  1. banking-knowledge-rag-assistant banking-knowledge-rag-assistant Public

    Production-style banking RAG assistant with FastAPI, Chroma, hybrid retrieval, query rewriting, reranking, citations, evaluation quality gates, Docker, CI, and web UI.

    Python

  2. banking-support-agent banking-support-agent Public

    FastAPI banking support agent with controlled tool use, confirmation-gated workflows, PostgreSQL persistence, Docker, CI, and AWS ECS/Fargate + RDS deployment proof.

    Python

  3. multi-provider-llm-starter-kit multi-provider-llm-starter-kit Public

    Reusable multi-provider LLM starter kit with Gemini, Groq, Mistral, streaming, structured JSON outputs, FastAPI, Docker, tests, and CI.

    Python

  4. databricks-ai-observability-lakehouse databricks-ai-observability-lakehouse Public

    Portfolio-scale Databricks lakehouse for AI support-agent observability using PySpark, Delta Lake, Bronze/Silver/Gold pipelines, data-quality controls, orchestration, and dashboards.

    Jupyter Notebook

  5. ga-feature-weighted-spectral-clustering ga-feature-weighted-spectral-clustering Public

    Genetic-algorithm feature weighting for spectral customer segmentation, comparing static and temporal RFM representations across multiple spectral graph constructions.

    Python