AI Automation Engineer Intern
Built HoBo Tracker from scratch (Next.js, Supabase) with a Capacitor Android app and native FCM push actions, used daily by the team to track ~90 properties (20 hotels, ~70 villas/homestays) through onboarding
I'm Archit Aggarwal. I build agents, RAG pipelines, voice AI and LLM products that ship, with real backends, real auth and real users.
Hiring? Read the 30-second summary or paste your job description into BYTE for a fit check.
FRIDAY wants to run
system.power("shutdown")This is how F.R.I.D.A.Y, my voice assistant, handles risky actions: a code-enforced gate, then an independent verifier that re-checks real system state.

Voice-first Windows assistant with 37 action modules: it controls the PC (apps, files, browser, screen/OCR), Gmail, Calendar, Spotify and your phone via FCM, with Hindi/Hinglish speech and a lip-synced 3D face. A code-enforced Sentinel gate holds risky actions until you confirm, an independent Verifier re-checks real state, and an emotion engine fuses text, voice and facial signals.
Built HoBo Tracker from scratch (Next.js, Supabase) with a Capacitor Android app and native FCM push actions, used daily by the team to track ~90 properties (20 hotels, ~70 villas/homestays) through onboarding
Applied AI techniques to data analytics workflows under the AICTE–IBM SkillsBuild program
Transformer + LSTM-Attention model for Hinglish sentiment analysis — 92.1% accuracy
Automated Power BI + SQL dashboards — reporting 65% faster
Sensor integration and autonomous flight-control software on real UAVs
Specialised in agentic AI, LLM engineering and full-stack AI products
Science and mathematics focus