Skip to content
View abhijitmanna912001's full-sized avatar
πŸ™‚
Available
πŸ™‚
Available

Block or report abhijitmanna912001

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
abhijitmanna912001/README.md

Typing SVG

πŸ‘‹ About Me

  • πŸ”­ Actively looking for Frontend / Backend / Full-Stack Developer roles (0–2 yrs), open to Delhi NCR & remote
  • πŸ—οΈ Building AuditFlow, a multi-agent AI system that reviews audit evidence, routes exceptions and produces review-ready workpapers
  • πŸ€– Built SignalScout AI, a B2B prospecting agent that validates data and fact-checks its own LLM-written outreach before anything is sent
  • πŸ’Ό Recently worked as a Content Management Analyst at WPP (Microsoft account) and as a freelance AI Trainer (RLHF) at Outlier
  • πŸ“š Currently learning advanced Next.js topics
  • 🌱 Comfortable across the stack: React, Next.js, TypeScript, Node.js, Express, PostgreSQL, Prisma, MongoDB

πŸ›  Tech Stack


πŸš€ Featured Project

SignalScout AI: B2B prospecting agent that checks its own claims (live beta, invite-only)

It validates and deduplicates company records before any paid enrichment, then verifies every number in LLM-written outreach against cited facts and suppresses any draft that fails.

  • Benchmarked baseline vs improved pipelines on 10 real records: drafts with citation violations fell from 4 to 0 and model cost dropped from $0.0272 to $0.0194, traced end to end with neatlogs
  • Shipped on Vercel with hashed invite codes, per-IP rate limits, record and call caps and a database kill switch, backed by 829 Vitest tests
  • Next.js, React, TypeScript, PostgreSQL, Tailwind CSS, Anthropic API, Crustdata API

πŸ“‚ Other Projects

Project Description
AuditFlow (live) Autonomous audit-evidence review with 5 AI agents, evidence-backed exception routing and review-ready workpapers
ContractCommander Multi-agent contract risk analysis (Commander, 5 specialist agents, Critic, Aggregator) that outputs a 0–100 risk score. React, Node/Express, Prisma, Neon Postgres, Docker

🀝 Connect with Me

GitHub LinkedIn LeetCode Codeforces


πŸ“Š GitHub Stats

Pinned Loading

  1. contractcommander contractcommander Public

    Multi-agent AI system that analyzes contracts for risk β€” 5 specialized agents + a critic produce a scored, evidence-backed report.

    TypeScript 1

  2. auditflow auditflow Public

    Autonomous audit-evidence review with 5 AI agents, evidence-backed exception routing, and review-ready workpapers.

    TypeScript 2

  3. mindstack-neathack mindstack-neathack Public

    SignalScout AI: a B2B prospecting agent that validates company data and checks its own claims before spending on enrichment or writing outreach. Evidence runs, neatlogs traces and an invite-only li…

    TypeScript 2