Skip to content
View adityasharmaaaaa's full-sized avatar
5 followers · 8 following

Achievements

Achievement: Pull SharkAchievement: YOLOAchievement: Quickdraw

Achievements

Achievement: Pull SharkAchievement: YOLOAchievement: Quickdraw

Block or report adityasharmaaaaa

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
adityasharmaaaaa/README.md

Hi, I'm Aditya Sharma

Final-year Engineering student at IIIT Bhopal. I build high-performance ML and backend systems with a focus on GPU computing, distributed architectures, LLM workflows, and low-level optimization.

What I've Built

  • cracked.c — ML library built from scratch in pure C and CUDA
  • CUDA-optimized RAG reranking pipeline with custom fused GPU kernels
  • Corrective RAG systems with self-evaluating retrieval and fallback workflows
  • Autonomous multi-agent systems using LangGraph and MCP
  • LLM evaluation and benchmarking infrastructure for coding agents
  • High-performance backend APIs, asynchronous pipelines, and data systems

Tech Stack

  • Languages & Systems C · C++ · CUDA · Python · Java · SQL · Linux
  • ML & GPU PyTorch · Triton · HuggingFace · vLLM · Nsight Compute/Systems
  • AI & Data LangGraph · MCP · RAGAS · DeepEval · ChromaDB · Qdrant
  • Infrastructure FastAPI · PostgreSQL · Redis · Celery · Docker · GitHub Actions

Achievements

Competitive Programmer

  • LeetCode Guardian (Top 0.08% globally), Codeforces Specialist, CodeChef 4-Star
  • Global Rank 168/28k+ in CodeChef Starters and Rank 195/30k+ in LeetCode Biweekly

Open Source Contributor

  • Merged a fix into PyTorch for an incorrect Triton kernel launch grid causing flaky GPU validation
  • Fixed a bug in LeetGPU's JAX compiler affecting GPU kernel compilation

Interests

ML Systems · GPU Computing · Distributed Systems · High-Performance Computing · LLM Infrastructure · Backend Engineering · Competitive Programming

Pinned Loading

  1. cuda-reranker cuda-reranker Public

    A hand-optimized CUDA attention kernel for RAG cross-encoder reranking, achieving ~25% lower attention latency than PyTorch SDPA on an NVIDIA T4.

    Python 1

  2. repo-trace repo-trace Public

    Corrective RAG assistant for exploring and debugging unfamiliar codebases, with retrieval grading, query refinement, reranking, web fallback, and citation guardrails.

    Python 1

  3. cracked.c cracked.c Public

    From scratch neural network & autograd engine in C, arena allocator, PCG PRNG, unity build, zero dependencies.

    C 1

  4. train-maxx train-maxx Public

    Python

  5. flash-attention flash-attention Public

    Cuda

  6. WikiRace-RL WikiRace-RL Public

    Python