Ahmed Bootaan

Open to Summer 2027 software engineering internships

Ahmed Bootaan

I build AI systems and developer tools.

Computer science student at the University of Minnesota, graduating May 2028. Lately that means retrieval agents, a code visualizer for algorithm practice, and simulators for trading systems.

  • LocationMinneapolis, MN
  • StudyingBS Computer Science, University of Minnesota
  • GraduatingMay 2028
  • Looking forSummer 2027 internship
Ahmed Bootaan in a black suit and tie

I like building tools that make hard things visible: what a retrieval agent actually looked at before it answered, what your code is doing line by line, how an order book reacts to a burst of trades.

On campus I am part of the Social Coding Club, the Blockchain Club, and the AI Club. Outside of class I am usually working through NeetCode problems or starting another side project.

  1. AlgoPlay

    2026

    A coding-practice app that animates what an algorithm is doing: pointers moving through an array, linked-list arrows turning around, a stack pushing and popping, a breadth-first search spreading through a graph.

    A recording of AlgoPlay: opening the graph breadth-first search problem, switching to the Visualizer tab, and pressing play as the search spreads through the graph node by node.

    Six problems, from Two Sum to graph BFS, each with an explanation, hints, a JavaScript editor, and test cases that run against your own code. This is the fourth rewrite; the first version instrumented arbitrary JavaScript and Python through Acorn and Pyodide.

    TypeScript / React / ViteLive demo Source
  2. Practical-Agents

    2026

    An agentic RAG system that picks its own document-search tools, cites what it used, and abstains when the retrieved evidence does not support an answer.

    On a small, hand-written set of 23 questions it answered 20 correctly and never cited a source that did not back up its answer. Most of the misses were retrieval, not reasoning: the right passage was never fetched. Tuning chunk size and adding a reranker fixed that on this set, though 23 questions is too few to call it solved. Runs on Azure or fully offline.

    Python / Azure OpenAI / Azure AI Search / FAISS / RAGSource
  3. permit-harness

    2026

    A Python library and CLI for AI-assisted review of municipal permit applications that a city could actually audit.

    Rule-based checks run before any model is called, every finding is tied to the exact passage it came from, high-risk findings are held for a human reviewer, and prompts are redacted before they are logged. It works with any model provider, or none, and has no runtime dependencies.

  4. mm-lab

    2026

    A market-microstructure lab in C++20: an order-matching engine checked against invariants, benchmarked on fixed, repeatable workloads.

    C++20 / CMakeSource
  5. lob-engine

    2026

    A vectorized limit-order-book simulator with microstructure features, a backtester that checks for look-ahead leaks, and a numba benchmark suite.

    Python / NumPy / NumbaSource
  6. url-shortener

    2026

    A URL shortener with Redis caching, JWT login, per-user rate limiting, and a Docker setup that runs the whole thing with one command.

    Java / Spring Boot / Redis / DockerSource
Languages
Python, Java, C++, TypeScript, JavaScript, Go, SQL
AI & retrieval
Azure OpenAI, Azure AI Search, FAISS, RAG, LLM evaluation, Prompt engineering
Web & data
React, Spring Boot, Streamlit, NumPy, Pandas, SciPy, Plotly
Tools & platforms
Git, Docker, Linux, PostgreSQL, Redis, Pytest, JUnit, Gradle
  • Applying for Summer 2027 software engineering internships.
  • Working through NeetCode problems; my submissions are public on GitHub.
  • Active in the Social Coding Club, the Blockchain Club, and the AI Club at the University of Minnesota.

Email is the best way to reach me. I am looking for a Summer 2027 software engineering internship, and I am always happy to talk about any of the projects above.

ahmedbootaan33@gmail.com