Skip to content
View AJOYSR's full-sized avatar
🎯
Focusing
🎯
Focusing

Highlights

  • Pro

Block or report AJOYSR

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
AJOYSR/README.md

Ajoy Sarker

Software Engineer | AI/ML Researcher

LinkedIn GitHub Email Portfolio Hugging Face


About Me

I am a Software Engineer at Brain Station 23 and an AI/ML Researcher. I work on reliable LLM systems — RAG pipelines, LangGraph agents, and guardrails that catch bad outputs before they reach a user.

My research sits at the intersection of trustworthy AI, LLM / RAG evaluation, and low-resource NLP.

  • Building agentic LLM systems and retrieval reliability tools in production
  • Open to collaboration on LLM agents, RAG evaluation, and multilingual NLP

Research Interests

Trustworthy AI Reliable LLMs Agent Guardrails Low-Resource NLP Model Evaluation Robust ML


Experience

Software Engineer · Brain Station 23

2024 – Present · Dhaka, Bangladesh

RAG systems Pipelines over structured knowledge bases (~92% retrieval accuracy, sub-200ms latency) with output validation
LLM agents LangGraph workflows with structured outputs and safety guardrails — 1st place, AI Hackathon 2025
Document AI Gemini Vision pipelines for schema-constrained extraction at 100K+ records/day

Technical Skills

Languages

Python TypeScript JavaScript SQL C++

ML / NLP

PyTorch Hugging Face Transformers scikit--learn

LLMs & Agents

LangGraph RAG Guardrails LoRA%20/%20PEFT

Systems

FastAPI PostgreSQL pgvector Docker AWS

Pinned Loading

  1. AI-Hackathon-2025 AI-Hackathon-2025 Public

    A custom chatbot built with NestJS, MongoDB, PostgreSQL with pgvector extension for vector similarity search. This project implements a sophisticated chatbot system with authentication, monitoring,…

    TypeScript 3 1

  2. NID-Parser-Bangladesh NID-Parser-Bangladesh Public

    A FastAPI-based OCR service that extracts structured information from National ID (NID) images using EasyOCR.

    Python 1 2

  3. BD-Nid-Bank-Parser-AI BD-Nid-Bank-Parser-AI Public

    Fast, accurate OCR extraction for document types using **Google Gemini 2.5 Flash Lite**. Optimized for speed with 1-2 second response times.

    TypeScript

  4. Abusive-language-detection Abusive-language-detection Public

    Jupyter Notebook