Headline

AI Engineer @ Dwell Properties | Agentic AI RAG & Backend Systems | LangGraph · FastMCP · Node.js · TypeScript | Ex-Yellow.ai (Series E) 🇮🇳

About

An AI agent that demos well is easy to build. One that works in production — routing decisions, calling tools, recovering from failures — is a systems problem. That's the work I do. Software Engineer at Yellow.ai (Series E) — 1.5 years building backend infrastructure and agentic AI systems inside a high-velocity enterprise SaaS environment. ▸ TicketOps LangGraph + Azure OpenAI + RAG + custom FastMCP tool server — built from scratch. Classifies IT tickets, calls diagnostic APIs, maintains workflow state, auto-generates structured resolutions. Full agentic pipeline. No prior template. No hand-holding. ▸ Edith Slack automation engine. Zero direction. Zero template. 12+ workflows · 1,000+ requests/month — approval routing, incident escalation, plan upgrades, ops automation. Built during my internship. Became the team's primary automation platform before I was promoted. ▸ The Incident Google Chat down for 10,000+ Royal Enfield enterprise users. No docs. Live production. Clock running. Built the recovery. It held. Beyond the headline work: → YmLogger — Pino + AsyncLocalStorage; resolved K8s pod-restart log loss, enabled distributed tracing across async chains → Work Portal — REST APIs, S3 flows, auth for 500+ daily users; AI feedback tracking 400+ users and 300+ monthly AI responses → GPT-4o RCA pipeline — automated validation of external RCA reports against compliance rules before sharing; cut manual review effort → GPT-4o semantic dedup — Slack + Google Sheets pipeline cutting duplicate feedback by 40% → 10+ RESTful APIs with x-api-key auth, RBAC, and per-route rate limiting Node.js · TypeScript · Python · PostgreSQL · Redis · LangGraph · FastMCP · RAG · Azure OpenAI · LangChain · Docker · Kubernetes → Backend engineering for APIs, workflows, and internal platforms → Agentic AI pipelines — tool use, memory, routing, and guardrails → Systems that hold at 3am, not just in demos MCA · NIT Bhopal · NIMCET AIR 360 · CGPA 8.9 Open to: Software Engineer · Backend Engineer · AI/Agentic AI Engineer Remote · Bengaluru · Gurugram · Noida · Hyderabad · Pune 📩 chahatgupta009@gmail.com · DMs open

Location

Agra

AI stack

codex

cursor

opencode

Work experience

Artificial Intelligence Engineer at Dwell Properties, Jul 2026 - Present · 2 mos, Noida

Artificial Intelligence Engineer at Daten Technology Solutions, Jul 2026 - Present · 2 mos, Noida

Software Engineer at Yellow.ai, Sep 2025 - Jul 2026 · 11 mos, ▸ Architected TicketOps — a LangGraph multi-agent pipeline integrating Azure OpenAI, RAG with contextual memory, and a custom FastMCP tool server; enables autonomous IT ticket classification, prerequisite checking, diagnostic tool calls, and structured resolution generation at enterprise scale ▸ Built the FastMCP Tool Server from scratch — exposing 10+ internal capabilities (OpenSearch log analysis, voice RAG, bot config APIs) so AI agents can safely act on enterprise systems instead of only generating text responses ▸ Led emergency incident resolution for a Google Chat outage affecting 10,000+ Royal Enfield enterprise users — diagnosed schema compatibility failures, migrated Cards v1 to CardsV2, and restored service within one business day with zero handoff documentation ▸ Built GPT-4o semantic deduplication pipeline (Slack + Google Sheets integration) cutting duplicate feedback by 40%; separately built GPT-4o RCA validation pipeline auto-verifying external RCA reports against compliance rules before team sharing ▸ Implemented YmLogger (Pino + AsyncLocalStorage) — resolved critical log-loss on Kubernetes pod restarts and enabled reliable distributed tracing across async chains in production ▸ Designed Work Portal features for 500+ daily users; built thumbs-up/down AI feedback system tracking quality signals across 400+ users and 300+ monthly AI responses — adopted as primary feedback mechanism across all support teams ▸ Architected 10+ RESTful APIs with x-api-key auth, RBAC, and per-route rate limiting; built AWS S3 presigned URL pipeline (generate → upload → fetch) with versioned v2 API for third-party consumers; reduced authenticated DB lookups by 50% via NextAuth JWT embedding

Software Engineer Intern at Yellow.ai, Jan 2025 - Aug 2025 · 8 mos, ▸ Proactively built Edith from scratch with no prior direction — a Slack automation engine running 12+ enterprise workflows (approval routing, incident escalation, plan upgrades, ops automation); scaled to 1,000+ requests/month and became the team's primary enterprise automation platform by end of internship ▸ Engineered end-to-end approval workflows using Slack API webhooks and event-driven architecture; automated plan upgrades and subscription management, reducing manual intervention by 70% ▸ Built Slack Incident Manager with slash commands, multi-step modals, formatted incident reports, and real-time PostgreSQL sync — tracking 100+ incidents across internal teams

Contributor at GirlScript Summer of Code, Oct 2024 - Nov 2024 · 2 mos, ✔️ Made impactful contributions to open-source projects, improving functionality and user experience. ✔️ Collaborated with talented mentors and developers worldwide, enhancing my coding skills and knowledge. ✔️ Worked with a diverse tech stack, including Postman, to streamline API testing and improve project efficiency. ✔️ Gained hands-on experience in problem-solving, teamwork, and building scalable solutions.

AI Software Developer at Soul AI, Jun 2024 - Oct 2024 · 5 mos, Remote, ▸ Contributed to AI model quality evaluation and training data pipelines for LLM fine-tuning workflows ▸ Structured and processed large-scale datasets to improve training signal quality; applied data quality frameworks to reduce noise in annotation corpora ▸ Collaborated with cross-functional AI teams on NLP model evaluation, conversational AI refinement, and task performance benchmarking

Education

Master of Computer Applications - MCA, Computer Science, Maulana Azad National Institute of Technology

Bachelor of Computer Application, Computer and Information Sciences and Support Services, Dr. B. R. Ambedkar University, Agra