Skip to content
View u7k4rs6's full-sized avatar
:shipit:
:shipit:
6 followers · 10 following

Achievements

Achievement: Pull Sharkx2Achievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: YOLO

Achievements

Achievement: Pull Sharkx2Achievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: YOLO

Highlights

  • Pro

Block or report u7k4rs6

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
u7k4rs6/README.md
Utkarsh Bahuguna

Second year CS at BITS Pilani × Scaler School of Technology.
I build AI agents, reinforcement learning environments, and security tooling.

LinkedIn Email Portfolio

Building

Project What it is
Yojaka Real-time AI debate platform. Multiple agents argue a motion in a structured reasoning format. Next.js.
Incident Response OpenEnv RL-compatible agent environment that simulates microservice failures with a diagnostic action mechanic. Deployed on HuggingFace Spaces.
PROCTOR.AI B2B platform that scores AI-assisted engineering skill in technical interviews. Model-agnostic scoring core.
ThreatSim Cybersecurity attack / defense simulator. Tech Excellence Award.

Selected work

  • Top 100 solo finalist, Meta × PyTorch Hackathon (built the incident-response RL environment end to end)
  • Open source contributor to Jaeger GenAI observability (jaegertracing/jaeger-ui #3887, LFX Mentorship)
  • Top 1% delegate, Japan Youth Summit 2025 (UNESCO affiliated), pitched an AI cybersecurity education model
  • Runner-up, Arweave Hackathon

Stack

Languages  TypeScript Python Go C++ Java Frameworks  Next.js React FastAPI Node Infra  Docker Postgres Redis GCP Focus  LLM agents reinforcement learning observability appsec

Pinned Loading

  1. MIRR MIRR Public

    MIRR is a simulation lab for training and evaluating heuristic and LLM agents to diagnose and recover microservice incidents, with a live Gradio demo for interactive incident response.

    Python

  2. Shadowbook Shadowbook Public

    A Rust limit order book matching engine, differentially fuzzed against an independent reference. The bugs were in the machinery checking the engine, not the engine

    Rust

  3. CAIRN CAIRN Public

    Self-hosted Git host on a from-scratch VCS engine in Java: content-addressable object store, packfiles with delta compression, three-way merge, and smart-HTTP transport, all cross-verified against …

    Java

  4. compute-elasticity compute-elasticity Public

    Pre-registered pilot characterizing how LLM reasoning accuracy scales with inference-time compute. Fits parametric curve families to per-problem response dynamics on GPQA Diamond.

    Python

  5. Tessera Tessera Public

    Decentralized, poisoning-resistant distribution for ML models and datasets. Publishers sign once, untrusted mirrors distribute, and any consumer verifies the exact signed bytes with full provenance.

    Python

  6. Flint Flint Public

    x86-64 kernel written from scratch in Rust and inline asm. Paging, frame allocator, kernel heap, preemptive scheduler, IDT/PIC interrupt path, and a shell. Runs under QEMU.

    Rust