Skip to content
View mblauberg's full-sized avatar
  • University of Queensland
  • Brisbane, QLD
  • 07:33 (UTC +10:00)
  • LinkedIn in/mblauberg

Block or report mblauberg

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.

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
mblauberg/README.md

Michael Blauberg

Final-year software engineering (Hons) student at the University of Queensland, Brisbane. I build AI/ML systems and the architecture around them — most recently agentic engineering: LLM agents in governed, reviewable multi-agent workflows, doing real product work.

Selected work

  • Provenant: portable equal-primary agent harness for Claude and Codex. Skills, review gates and an agent fabric for a governed agentic SDLC.
  • RAG-as-a-Service: document Q&A service with hybrid retrieval, reranking and source citations. FastAPI, React, containerised, with Kubernetes manifests.
  • GlassChain: read-only blockchain wallet intelligence system from UQ's software architecture course. TypeScript monorepo, CI, load testing, deployed on AWS. My focus was the microkernel plugin architecture. (Team project, private repo.)
  • Nova Edge: underwater glider simulator in Unreal Engine (C++), built for Flying Fish Technologies. I built the physics systems: hydrodynamics, PID-controlled attitude and spring-damper cable towing. (Team project, private repo.)
  • My Boardgame Collection: full-stack product with React 19, TypeScript, Supabase and passkey (WebAuthn) authentication — live on Vercel.
  • Beast Breakout: 19 merged pull requests into a 40+ student Java/libGDX game, covering NPC and boss AI, effects, refactoring and tests. (Team project; my contribution only.)

I recently completed my honours thesis on ML-based scam detection. The work is private for now.

LinkedIn

Pinned Loading

  1. provenant provenant Public

    Portable equal-primary Claude and Codex agent harness for a governed agentic SDLC

    Python 1

  2. rag-as-a-service rag-as-a-service Public

    Retrieval-augmented generation service: upload documents, ask questions in natural language, get answers with source citations. FastAPI backend, React frontend, containerised.

    Python 1

  3. my-boardgame-collection my-boardgame-collection Public

    Board-game collection tracker: shared catalogue, personal library and play recommendations. Vite, React 19, TypeScript and Supabase, deployed on Vercel.

    TypeScript

  4. brain-dump-app brain-dump-app Public

    Turns freeform notes into structured tasks, habits and calendar events using an LLM. React 19, TypeScript and Zustand, with pluggable OpenAI, Claude or Gemini backends.

    TypeScript

  5. table-tennis-ml-classification table-tennis-ml-classification Public

    IMU sensor-data classification of table-tennis strokes with group-aware train/test splits to prevent player leakage. scikit-learn + LightGBM (UQ COMP4702).

    Python

  6. dqn-pytorch-implementation dqn-pytorch-implementation Public

    Deep Q-Network reinforcement learning agent in PyTorch on Gymnasium environments (UQ COMP3702).

    Python