Skip to content
View danielbisar's full-sized avatar

Organizations

@sprintworx-gmbh

Block or report danielbisar

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
danielbisar/README.md

Daniel Bisar

Senior Software Engineer — .NET, AI-Assisted Engineering & Developer Tooling

I’m a software engineer with 15+ years of professional experience building and evolving software systems, primarily with C#/.NET. My work combines hands-on implementation, architecture, debugging, technical leadership, and making complex codebases easier to understand and change.

I’m particularly interested in AI-assisted software engineering: using coding agents as engineering tools while keeping requirements, architecture, verification, testing, and technical judgment explicit.

Current focus

  • C# / .NET and long-lived software systems
  • AI-assisted engineering and coding-agent workflows
  • Developer tooling and engineering productivity
  • Architecture, codebase analysis, testing, and maintainability
  • Python / PyTorch experimentation
  • C++ / Linux / native and graphics integration

Selected projects

Experimental C# dependency explorer for codebase reconnaissance. It combines Roslyn syntax analysis with an interactive WebGL graph, with an explicitly heuristic analysis contract, automated regression coverage, and CI.

Small-language-model experiments in PyTorch focused on correctness and measurement: sparse Mixture-of-Experts routing, tokenization, EOS-aware evaluation, checkpoint reliability, reproducible experiment provenance, and quality/runtime/memory trade-offs.

Local non-root Docker and Podman environments for Codex, Claude Code, and Pi, with persistent state, version-aware image selection, explicit isolation limits, and regression-tested tooling.

Historical employer-supported C# UI framework for embedded-style applications. It separates control behavior from rendering to enable headless interaction testing and includes substantial SkiaSharp/OpenGL/native integration. Published with SprintWORX attribution and permission.

Bounded Linux/GLX rendering-interposition experiment using LD_PRELOAD, framebuffer capture, OpenGL state preservation, and a shader-based zoom inset. Kept intentionally small and explicit about its supported scope.

Open-source contributions

Contributed function-tool support and restorable conversation history to the .NET SDK's direct Codex OAuth integration. The change added streamed tool-call aggregation, host-controlled tool execution, automatic continuation with tool outputs, and typed history restoration that preserves function call IDs across process restarts.

Professional background

  • Senior IT Consultant / Technical Lead at msg systems ag
  • Hands-on development and technical leadership in business-critical software
  • Earlier experience leading a software-development department and an international development team
  • Domains include finance, logistics, embedded/IoT, and business software

Engineering approach

I prefer roles where architecture stays close to implementation. I like working directly with difficult systems: understanding existing behavior, reducing uncertainty, improving testability, debugging across abstraction boundaries, and building tools that make subsequent engineering work easier.

With AI-assisted development, I use agents aggressively for implementation and exploration, but treat generated code as untrusted until it is reviewed, tested, measured, and reconciled with the actual system behavior.

Links

Pinned Loading

  1. agent-sandbox agent-sandbox Public

    Local, non-root Docker and Podman environments for Codex, Claude Code, and Pi, with persistent state, version-aware image selection, and explicit isolation limits.

    Shell 2

  2. Medja Medja Public

    Historical C# UI framework for embedded-style applications, with headless control tests and SkiaSharp/OpenGL rendering.

    C# 2 1

  3. code-map code-map Public

    Experimental C# dependency explorer with Roslyn-based scanning and interactive WebGL graphs.

    C#

  4. tinylm tinylm Public

    TinyLM is a small language-model implementation in PyTorch for exploring model architecture, tokenization, training efficiency, and local-GPU constraints

    Python