AI & Machine Learning · Knowledge Graphs

Building intelligent systems that are
trustworthy by design.

I'm Mihai Mehedint — an AI and machine learning research scientist / engineer. I work on knowledge graphs, semantic search, and applied ML, with a strong bias toward systems that are secure, explainable, and built to last.

View my work Get in touch

About

Mission

My mission is to help turn scattered data into structured, connected knowledge — and to build AI systems people can actually trust.

I believe the next leap in machine learning comes not just from bigger models, but from better structure: knowledge graphs that give models memory and context, semantic search that understands meaning rather than keywords, and pipelines that are transparent about how they reach an answer.

I approach every system as an engineer first and a security-minded one second — favouring solutions that are explainable, privacy-respecting, and resilient. Good AI should be measurable, auditable, and honest about its limits. That principle guides how I design, build, and ship.

What I do

Focus areas

Knowledge graphs

Designing and populating graphs that give models structured memory — entities, relationships, and context that make answers grounded and traceable.

Semantic & vector search

Retrieval that understands meaning — embeddings, vector databases, and RAG pipelines that return the right information, not just matching words.

Applied machine learning

From data to deployed model — NLP, LLM integration, and ML systems built to be evaluated, monitored, and improved responsibly in production.

Toolbox

Skills & technologies

A snapshot of what I work with. This will grow over time.

  • Python
  • Machine Learning
  • Deep Learning
  • NLP
  • LLMs & RAG
  • Knowledge Graphs
  • Graph Databases
  • RDF / SPARQL
  • Vector Search
  • Embeddings
  • Data Engineering
  • MLOps

Selected work

Projects & highlights

A place for my projects and accomplishments — expanding as I publish more.

Project title

AI · Knowledge graphs

Short description of the project: the problem, your approach, and the outcome. Add links to a write-up, repo, or demo.

Project title

Machine learning · NLP

Short description of the project: the problem, your approach, and the outcome. Add links to a write-up, repo, or demo.

Off the clock

Interests & free time

Beyond the terminal — a few things that keep me curious.

Reading & ideas

Books and essays across science, technology, and philosophy.

Outdoors

Time away from screens — walking, travel, and the occasional adventure.

Building things

Side projects and experiments, from tinkering with hardware to design.

Contact

Get in touch

Open to collaboration, interesting problems, and good conversations.