Skip to content
View danielsolo707's full-sized avatar
🎯
Focusing
🎯
Focusing
1 follower · 1 following

Highlights

  • Pro

Block or report danielsolo707

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
danielsolo707/README.md

Hi, I'm Daniel Soleimany

ML engineer & builder — I train models, design agent pipelines, and ship usable products.

Focus: deep learning (audio / multimodal) · reinforcement learning · tool-using LLM agents · full-stack delivery

Pinned projects (start here)

Project What you'll find Proof
BirdCLEF Multi-label bird sound SED — EfficientNet-B0 + attention Val macro ROC-AUC 0.8529, weights + train/eval/infer
BipedalWalker-PPO Continuous control with PPO (SB3, 1M steps) Checkpoints + training progression video
AlphaMath Offline math agent: Qwen2.5-Math writes Python → sandbox → self-repair → majority vote AIMO-style constraints, no external APIs
studio2.0 Personal site & creative portfolio (Next.js, Three.js) Live: danielsoleimani.ir

Also shipping

Stack

Python · PyTorch · timm · Transformers · Stable-Baselines3 · Gymnasium · scikit-learn · SymPy · Next.js · TypeScript · Supabase

How I work

  1. Specify goal, constraints, and metrics
  2. Build (including AI as a force multiplier where it helps)
  3. Verify with tests, held-out metrics, and honest limitations

Open to ML engineering, applied AI, and roles where models meet production.

Web: danielsoleimani.ir · GitHub: danielsolo707

Pinned Loading

  1. AlphaMath AlphaMath Public

    ALPHA-MATH: offline System-2 math agent — Qwen2.5-Math writes Python/SymPy, sandbox executes, self-repair + majority vote (AIMO-style, no external APIs)

    Python

  2. BipedalWalker-PPO BipedalWalker-PPO Public

    Deep RL: PPO agent on Gymnasium BipedalWalker-v3 with training progression videos (1M steps, SB3)

    Python

  3. BirdCLEF BirdCLEF Public

    BirdCLEF+ 2025 multi-label bird sound event detection — EfficientNet-B0 SED + attention pooling (val AUC 0.8529)

    Python

  4. studio2.0 studio2.0 Public

    TypeScript

  5. gesture-canvas-museum gesture-canvas-museum Public

    An interactive digital art playground where mouse and hand gestures drive drawing, color changes, undo, and canvas clearing. Designed with a performant Three.js rendering engine and MediaPipe-based…

    JavaScript