Trilochan
ProjectsBlogsResearchRésumé
ProjectsBlogsResearchRésuméContact
CONTACT

Let's work together

parnishklpo@gmail.com ↗
LinkedIn ↗GitHub ↗WhatsApp ↗
© 2026 Trilochan Sharma·Built with Next.js·Kathmandu, NPBack to top ↑
JOB AGENT
Home/Explore/AI/AGENTIC-AI

Project

JOB AGENT

"A full-stack agentic pipeline that scrapes 5 job boards daily, scores listings with a structured AI rubric, rewrites your resume per role, and submits — with two mandatory human gates hardwired in. Zero unreviewed submissions.

"A full-stack agentic pipeline that scrapes 5 job boards daily, scores listings with a structured AI rubric, rewrites your resume per role, and submits — with two mandatory human gates hardwired in. Zero unreviewed…

LangGraphFastAPINext.js 14PostgreSQLpgvectorCelery+4
View FolderRepositoryDetails
Statusshipped
Categoryai
Stack10 tools
Updated2026-03-30T05:24:49.225636+00:00
Timeline6 steps

Overview

Key Features

  • ai
  • automation
  • langgraph

Project Timeline

Step 1
Project ideation & architecture design

esigned the 5-stage pipeline architecture, chose tech stack, planned LangGraph agent graph and RL feedback loop.

Step 2
Core scraping pipeline built

Integrated JobSpy and Playwright to scrape LinkedIn, Indeed, and Glassdoor. Built deduplication and PostgreSQL storage layer.

Step 3
Semantic job scoring engine

Implemented pgvector embeddings for job-profile matching. Built LangGraph scoring agent with conditional routing.

Step 4
DPO resume generation model

Fine-tuned resume generation model using HuggingFace TRL with DPO. Implemented ATS keyword mirroring per job description.

Step 5
Browser automation agent

Built Playwright agent to autonomously navigate job portals, fill forms, upload resumes and submit applications via FastMCP.

Step 6
Dashboard MVP

mplementing nightly DPO retraining pipeline using application outcome signals (rejection, view, interview) as reward feedback.

More Projects

Intelligent-Product-Pitch-Recommendation-System.01
Shipped

Intelligent-Product-Pitch-Recommendation-System.

An AI-powered system that generates personalized product pitches by analyzing customer context and intent. It helps businesses move from…

Next.jsTypeScriptSupabasePostgreSQL+2
Repo ↗
customer-churn-prediction02

customer-churn-prediction

A machine learning project that predicts customer churn by analyzing historical customer data. It uses data preprocessing, feature…

PythonpandasNumPyscikit-learn+5
Live ↗Repo ↗
cms-mcp03
Shipped

cms-mcp

An MCP server that gives Claude full natural-language control over any REST CMS — 32 auto-generated tools, human approval gate, policy…

TypeScript 5.8Node.jsMCP SDKZod+3
Repo ↗