Anubhav Paul@anubhav465
2
Events attended
2
Submissions made
India
1 year of experience
About me
I’m Anubhav Paul, a student passionate about building a career in Artificial Intelligence and Machine Learning, with a strong focus on solving real-world problems using modern AI systems. I work on developing systems like recommendation engines, chatbots, and RAG-based applications that help improve user experience and decision-making. My approach involves using LLMs, Retrieval-Augmented Generation (RAG), advanced RAG techniques, and machine learning to create scalable and efficient solutions. With my academic background and continuous hands-on learning, I’ve developed strengths in critical thinking and problem-solving, allowing me to approach challenges with a structured and analytical mindset. I focus on turning ideas into working systems that create value. I believe in learning by building and experimenting, and I’m constantly exploring new ways to apply AI in impactful domains. Currently, I’m open to internships, research opportunities, and entry-level roles in AI/ML, where I can contribute, learn, and grow in a fast-paced environment.
Socials
🤝 Top Collaborators
.png&w=640&q=75)
.png&w=640&q=75)
🤓 Latest Submissions
.png&w=828&q=75)
Neural Analytics
An AI-powered analytics platform that lets you upload CSV files and ask natural language questions. Supports multi-table querying with automatic JOINs, semantic schema search, and AI-generated insights with interactive visualizations. Drag-and-drop CSV upload — Upload up to 5 files (40 MB each) instantly Multi-table NL querying — Ask questions across multiple datasets with automatic JOINs Zero SQL knowledge required — LLM generates correct queries from plain English Semantic schema search — FAISS embeddings find relevant columns by semantic meaning Auto-fix on failure — LLM self-corrects failing queries in real-time Interactive visualizations — Auto-selects best chart type (bar, line, pie, histogram) AI summaries — 3-bullet findings generated from query results Data cleaning report — Shows nulls normalized, duplicates removed, type coercions
19 May 2026
AIRA - Autonomous Incident Response Agent
AIRA (Autonomous Incident Response Agent) is an AI-powered incident response automation platform that revolutionizes how teams handle production incidents. Built with FastAPI, React, and LangGraph, AIRA receives incidents from monitoring tools like Datadog or Prometheus and executes a sophisticated 5-phase workflow powered by Groq's Llama 3.3 70B LLM. The system automatically classifies incident severity (P0-P3) in under 5 seconds, fetches relevant code context via MCP servers, analyzes similar historical incidents using Redis-based similarity search, generates git-style patches with unit tests, and calculates a composite confidence score. Based on this score, AIRA either automatically creates GitHub pull requests (>85% confidence) or requests human approval via Slack. With enterprise-grade security features including blocked paths for critical files, complete audit trails, and mandatory human oversight for P0 incidents, AIRA achieves 80% faster resolution times while maintaining safety. The real-time WebSocket dashboard provides full transparency into agent decisions, and the system continuously learns from each incident to improve future responses. AIRA reduces incident resolution time from hours to seconds, saves $4,000/hour in downtime costs, and cuts on-call burden by 60%—all while maintaining 95% accuracy in severity classification.
17 May 2026
