AB

Aviral Bagjani

AI Engineering Intern · Full-Stack ML/LLM Developer

Kanpur, Uttar Pradesh, India

@aviral_bag

0 followers

🎓 B.Tech Electronics and Communication Engineering at Indian Institute of Information Technology Bhagalpur · Graduating 2026

About

B.Tech ECE student at IIIT Bhagalpur with hands-on experience building AI automation, RAG, and multi-agent systems using modern LLM stacks. Strong focus on Python, prompt engineering, retrieval systems, and production-oriented testing/observability.

Experience

  • AI Engineering Intern
    Zeepty
    Apr 2025 – Jul 2025

    Built AI-driven automation creator-commerce workflows, applying semantic creator/product matching and personalized LLM outreach across 500+ combinations, improving relevant-match retrieval by 25%. Improved LLM-generated creator outreach by refining prompts and retrieved context, increasing response relevance by 15% across 20+ representative scenarios. Engineered and tested AI workflow components using Python and Git/GitHub, with GitHub Actions for automated CI checks across multiple development builds; iteratively tracked and resolved matching and LLM-generation issues through 50+ testing iterations.

Education

Skills

Tools / apps / platforms

3 to 6 months 3 to 6 months A few months of practice with it.
Git GitHub GitHub Actions
<3 months <3 months Just getting started - under three months. This is also what shows when a level has not been set.
n8n Groq Neo4j Redis Docker OpenAI QDrant Vercel Kubernetes PostgreSQL Google Gmail Hugging Face Microsoft Excel Amazon Web Services AWS

Projects

  • Python, LangChain, LangGraph, Qdrant Cloud, Redis, PostgreSQL, Groq, NeMo Guardrails, LangSmith, FastAPI, Streamlit

    Built and deployed an agentic RAG system with hybrid retrieval and citation-grounded generation. Added caching, rate limiting, checkpoints, guardrails, and regression evaluation to improve latency, scalability, reliability, and cost efficiency.

  • Python, LangGraph, LangChain, FastAPI, PostgreSQL, MCP, Gmail API, Tavily, Groq, LiteLLM, NeMo Guardrails, LangSmith, Streamlit, DeepEval

    Designed a planner-executor multi-agent orchestration system with research, blog generation, and email workflows, reducing a 15-minute research-to-draft process to under 2 minutes. Integrated external tool ecosystems through MCP with human-in-the-loop approval, and hardened the platform with guardrails, JWT authentication, and evaluation metrics.

Courses & certifications

  • Communication · Indian Institute of Information Technology Bhagalpur · 2026
  • OOP · Indian Institute of Information Technology Bhagalpur · 2026
  • Computer Networks · Indian Institute of Information Technology Bhagalpur · 2026
  • Operating Systems · Indian Institute of Information Technology Bhagalpur · 2026
  • DBMS · Indian Institute of Information Technology Bhagalpur · 2026
  • Data Structures & Algorithms · Indian Institute of Information Technology Bhagalpur · 2026

🏆 Achievements & awards

  • KNIGHT Badge on LeetCode · 2025

    Achieved KNIGHT Badge (rating 1784) and solved 500+ coding questions.

  • JEE MAIN 2022 All India Rank 52,750 · 2022

    Scored 95.44 percentile out of 1.2 million candidates.

Explore the Broxer community

🤖
Online · instant AI help
Broxer