FazBrowse GitHub Viewer | Trending |
URL:
| Home
Tools: [Original HTTPS Page]

abhishekacharya200 (Abhishek Acharya) · GitHub

Overview Repositories 43 Projects Packages Stars 11
🏠
Working from home
🏠
Working from home
18 followers · 18 following

Achievements

x2

Achievements

x2

Highlights

  • Pro

Block or report abhishekacharya200

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.

Close all issues, pull requests, and discussions opened by this user Content in all repositories owned by your account will be closed.
Add an optional note
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
abhishekacharya200/README.md

Abhishek Acharya

AI engineer in Kathmandu, Nepal. I build retrieval (RAG) and agent systems, and I measure how well they work. Open to remote roles.

Portfolio · LinkedIn · Email

What I'm working on

  • AI and Automation Intern at K&A Power: a document search service over about 50,000 company files, a daily tender tracker for ADB, World Bank and AIIB notices, and invoice matching for 1000+ documents a month.
  • Open source: fixing bugs in Python and AI projects, currently focused on LiteLLM and PyTorch Lightning.

Featured projects

Project What it is Proof
Production RAG (private, code on request) Document search and QA: hybrid dense and BM25 search in Qdrant, LLM reranking, per document access control, cited answers nDCG@10 0.795 on the public BEIR SciFact benchmark (dense only: 0.646, BM25: 0.667)
Fixgate Agent that reproduces a GitHub issue in a sandbox, writes and checks a fix, and opens a PR only after a person approves the exact patch Resolved 22 of 31 SWE-bench Verified issues (71%, partial run) at about $0.02 each
vault2okf Converts Obsidian vaults into Open Knowledge Format bundles for AI agents, then validates and repairs them Published on PyPI, with a GitHub Action and pre-commit hook
RAG Gateway Local server that speaks the OpenAI, Anthropic and Gemini APIs, for building and testing RAG apps Used to run the Production RAG benchmark through local OpenAI compatible endpoints

Open source contributions (merged)

All my merged pull requests

Tools I use

  • AI: RAG, hybrid search, reranking, LLM evaluation, agents and tool calling, OpenAI, Claude and Gemini APIs, Ollama
  • Search and data: Qdrant, FAISS, ChromaDB, PostgreSQL, Redis
  • Engineering: Python, FastAPI, Django REST Framework, TypeScript, Docker, GitHub Actions

Pinned Loading

  1. Fixgate Fixgate Public

    GitHub issue agent with guardrails: triage, reproduce in a sandbox, fix, verify, and open a draft PR only after human approval.

    Python

  2. vault2okf vault2okf Public

    Convert Obsidian vaults to Open Knowledge Format (OKF) bundles, then validate and repair any OKF bundle. Python CLI, GitHub Action and pre-commit hook.

    Python 1

  3. ntcbot ntcbot Public

    this is the ntcchatbot using the rag , api of nepal telecom and the ollama based trained model

    Python 2

  4. Aarambha-AI-Customer-Service-Chatbot Aarambha-AI-Customer-Service-Chatbot Public

    An intelligent customer service assistant powered by Retrieval Augmented Generation (RAG), Gemini LLM, and real-time intent classification. Built with Django REST Framework and React.

    Python 1

  5. RAG-Gateway RAG-Gateway Public

    The Anthropic, OpenAI and Gemini APIs on your machine, answered by the Claude Code, Codex and Antigravity CLIs you're already logged in to

    Python 1

Back | FazBrowse Home | New Git URL