I specialize in designing autonomous multi-agent architectures, enterprise-grade RAG pipelines, and generative media workflows.
- 💼 By Day: Building internal AI automation pipelines as an Agentic AI Developer Intern at JK Papers.
- 🚀 By Night: Co-leading technical strategy and building GenAI video pipelines as a Tech Partner / CTO at Veldon Lab.
- 🎓 Pursuing a B.Tech in Computer Science (AIML) at Dronacharya College of Engineering (2023-2027).
- 🏆 Achievements: GATE Data Science & AI (DA) Qualified (AIR 7528) | Smart.AI Hackathon Winner | SIH 2024 Finalist.
- 🧠 Currently scaling: Multi-agent LLM orchestration (CrewAI, LangGraph), deep prompt sequencing (Higgsfield AI), and custom ComfyUI workflows.
- Specialties: Agentic AI (CrewAI, LangGraph), ComfyUI, RAG (LangChain), Vector DBs (ChromaDB, Pinecone), Prompt Engineering.
- Infrastructure: RESTful APIs, Microservices, MLOps, Hugging Face Spaces, PostgreSQL.
- Technical Partner & CTO | Veldon Lab (Remote, US)
- Architecting multi-agent logic frameworks for automated UI/UX generation, boosting code accuracy by 40%.
- Engineering complex generative video workflows using ComfyUI and optimizing prompts for Higgsfield AI models.
- Agentic AI Developer Intern | JK Papers (Faridabad, India)
- Designing enterprise-grade Agentic AI workflows using LangGraph and CrewAI to automate core internal business logic.
- React Developer | Brandfortip (Noida, India)
- Authored core front-end components, achieving a 15% reduction in API response times.
- ML Developer Intern | JCB India Headquarters (Faridabad, India)
- Deployed machine learning models via REST APIs for Supply Chain optimization, achieving 94% accuracy in demand forecasting.
- 🧠 NeuroDecode - AI Agentic Platform: Built an AI system translating complex CV/NLP architecture diagrams into executable PyTorch code. Integrated a multi-agent workflow powered by Gemini 2.5 Flash and RAG to process diagrams in under 15 seconds.
- 🛡️ Multilingual AI System for Online Radicalization Detection: Constructed a two-step classification pipeline achieving 99% accuracy across English, Urdu, and Hinglish streams using Trie data structures and an RNN-LSTM model.
- ☁️ Microsoft Certified: Azure Data Fundamentals (DP-900)
- 🤖 Databricks: Generative AI Agent Fundamentals
- 🥇 NPTEL (IIT Ropar): Deep Learning (Gold Elite - Top 5%)
- 🥇 NPTEL (IIT Guwahati): Neural Networks for CV & NLP (Gold Elite)
