Skip to content
View yash194's full-sized avatar
1 follower · 2 following

Block or report yash194

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 supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
yash194/Readme.md

Yash Aggarwal

Building multi-agent systems, RL swarms, and practical AI products.

GitHub LinkedIn Email

🔥 What I’m focused on

  • Multi-Agent RL + Swarm Intelligence (coverage, search, coordination, energy-aware policies)
  • Agentic systems (debate + memory + distillation → “reasoning that improves over time”)
  • Applied AI products (automation, NLP classification, developer tooling)

🚀 Flagship Projects (Proof > Claims)

🛰️ GASAMC — UAV Swarm RL Framework

Graph-attention + Soft Actor-Critic + multi-critic training for cooperative drone missions

  • Coverage/search phases, custom metrics (coverage, redundancy, energy-per-area, efficiency)
  • Designed for research-grade ablations + paper-ready plots

🤖 Automation Products (n8n + Next.js + Supabase)

  • Classroom/assignment automation pipelines
  • YouTube script generator (web + scraping + structured outputs)

💬 Risk Signal Extractor (NLP)

Classifies support messages into categories (fraud/chargeback/legal threat/refund…) using classical ML / lightweight DL (no transformers)

🛠️ Tech Stack (what I actually use)

Core: Python • Pytorch • TensorFlow • NumPy • Pandas
ML/RL: Optax • Gym/Gymnasium • RL tooling
Backend: FastAPI • Node.js • REST APIs
Frontend: Next.js • React • Tailwind
Automation: n8n • Webhooks • Supabase
Dev: Git • Docker • Linux • CI (GitHub Actions)

📌 What I value (how I work)

  • Research-grade clarity: clean experiments, ablations, reproducibility
  • Engineering discipline: tests, readable diffs, minimal magic
  • Systems thinking: metrics, failure modes, memory/latency constraints
  • Ship mindset: demos that real users can try

📊 GitHub Stats

🤝 Let’s collaborate

If you’re working on multi-agent RL, agent memory/distillation, JAX tooling, or AI automation products — I’m open to collaborations.

Popular repositories Loading

  1. trex-runner-game trex-runner-game Public

    It appears when the internet is not available.

    JavaScript

  2. algorithm algorithm Public

    creating the collidal algorithm

    JavaScript

  3. banana-grab banana-grab Public

    a moving monkey will try to eat bananas and there are so many obstacles stopping him.

    JavaScript

  4. bounce-algorithm bounce-algorithm Public

    how to make the sprite bounce.

    JavaScript

  5. car-relaibility-tester car-relaibility-tester Public

    this will test the deformation of the car at different velocities.

    JavaScript

  6. car-relaibility-test car-relaibility-test Public

    it will test the damage on the car.

    JavaScript