Final year B.Tech AI student at Gautam Buddha University, building multi-agent automation workflows — mostly in n8n, orchestrated with Groq/LLaMA 3.3 70B.
I learn by building: I take an existing tool or partial workflow and extend it until it actually does something useful, then debug the parts that break.
- AI Ad Creative Agent (in progress) — a multi-modal pipeline that turns a single product prompt into a full video ad: copy → image → video → voiceover → final render. Image generation works end-to-end; video and text-overlay stages are still being debugged.
- Multi-Agent Social Media Content Generator — a 3-agent (Researcher → Writer → Editor) n8n pipeline that turns product inputs into Instagram captions, hashtags, and CTAs. Deployed on Railway via Docker. Live demo
- Agentic AI Web Search Agent — a ReAct-style agent with real-time web search (Tavily), a calculator tool, and session memory for multi-turn conversations. Live demo
n8n Groq API Docker Railway Tavily API Stability AI Google Sheets API Python
- LinkedIn: linkedin.com/in/anshikabharti
- Email: anshibharti64@gmail.com
Currently looking for an Agentic AI / LLM engineering internship.