🔭 I build AI and machine learning systems with attention to correctness, reliability, and practical constraints.
👯 I am open to collaborating on well-scoped, engineering-focused AI/ML projects and research-oriented systems.
🌱 I am deepening my understanding of generative AI, agentic workflows, and ML system design.
💬 Ask me about Python, machine learning fundamentals, NLP pipelines, model evaluation, and system-level tradeoffs.
📫 Reach me at brijeshrakhasiya.aiml@gmail.com or LinkedIn
⚡ I value steady progress through problem-solving and disciplined iteration.
- Surat, Gujarat, India
- in/brijesh-rakhasiya-bba776289
Highlights
- Pro
Pinned Loading
-
Deep-Research-Agent-From-Scratch
Deep-Research-Agent-From-Scratch PublicMulti-agent AI research system with LangGraph, FastAPI & RAG. Parallel research orchestration inspired by OpenAI Deep Research.
-
Agentic-Doc-Search-RAG
Agentic-Doc-Search-RAG PublicSelf-corrective Agentic RAG with LangGraph - eliminates hallucinations through intelligent relevance grading before answering. Features Streamlit UI, MCP server integration & multi-turn memory.
Python
-
AI-Customer-Review-Intelligence-System
AI-Customer-Review-Intelligence-System PublicProduction NLP platform: DistilBERT sentiment analysis (95% acc) + BERTopic issue clustering + A/B testing + Streamlit dashboard. Turn 200K Amazon reviews into business insights.
Jupyter Notebook
-
Agentic-Blog-Generator
Agentic-Blog-Generator PublicMulti-agent AI blog generator with LangGraph - creates SEO content with automatic translation to Hindi & French
Python
-
Agentic-Chatbot
Agentic-Chatbot PublicProduction-ready multi-agent AI system with LangGraph - featuring chatbots with real-time web search & automated AI news aggregation
Python
-
30-Days-Machine-Learning-Project
30-Days-Machine-Learning-Project PublicWelcome to my 30 Days of Machine Learning journey! 🎯 Over the past month, I committed to building 30 real-world, end-to-end ML projects—one project each day—to sharpen my skills, stay consistent, a…
If the problem persists, check the GitHub status page or contact support.


