Samarth S Shetty
I build AI agents, RAG pipelines & backends.
Ready for work
Samarth is ready for their next project!
Job seekers lose time and money to fake listings
every day. This agent catches them before they do.
Built a LangGraph-based AI agent that takes any
job posting URL or description, aggregates
intelligence from Reddit threads, Glassdoor
reviews, Google results, and company pages, then
analyzes red flags and returns a clear
verdict — scam, suspicious, or legitimate.
Uses Tavily for live web search, LangGraph for
multi-step reasoning, and FastAPI for the backend.
Deployed and live.
Built to show what a real agentic pipeline looks
like in production — not a chatbot, but a
reasoning system that investigates and decides.
1
7
Video creators waste hours manually transcribing
and captioning their content. CAPGEN eliminates
that entirely.
Built a full-stack AI web app that takes any video
upload, extracts the audio, transcribes it using
Groq Whisper, and generates accurate subtitles —
with the user choosing caption position (top,
middle, or bottom).
Live on Render with real users. Zero downtime
since launch.
Built for YouTube Shorts creators who need fast,
accurate captions without expensive editing
software.
1
18
Agricultural Disease Detection System
Description: Built a ResNet-50 based CNN model to detect and classify diseases in crops from images. Achieved high accuracy on agricultural datasets. Research was published as an IEEE paper and resulted in an approved patent in 2024.
Skills: Python, Computer Vision, Deep Learning
Tools: ResNet-50, OpenCV, NumPy
0
41
AI Resume Screener
Description: Built an AI-powered resume screening tool that ranks candidates against a job description. Users upload multiple PDFs, and the app analyses each using OpenAI API and returns ranked candidates with match scores and reasoning. Built with Python and FastAPI.
Skills: Python, OpenAI API, Prompt Engineering
Tools: FastAPI, Docker
0
50