Call Me Now
Call Me — Open to Opportunities
Available for new opportunities

Punit
Gautam

Staff Engineer · Python Backend · AI Systems

7+ years shipping high-scale distributed systems and production AI at Everest Tech, Data Works, Mirafra, and Walmart. I architect the backend layer that makes AI products — agentic pipelines built on LangChain / LangGraph — actually work in production.

Bangalore, India — open to remote
hiremeforse@gmail.com
+91 8880 144 666
Get In Touch Partner With Me ↓ resume.pdf
AI Agents
LangChain
LangGraph
AWS
Punit Gautam
Python FastAPI Django AWS GCP Docker PostgreSQL MongoDB GraphQL GitHub Actions LangChain LangGraph MCP CrewAI HuggingFace PySpark Jenkins ◆ Python FastAPI Django AWS GCP Docker PostgreSQL MongoDB GraphQL GitHub Actions LangChain LangGraph MCP CrewAI HuggingFace PySpark Jenkins ◆
99.99%
System Uptime
20K+
Requests / min
40%
AI Accuracy Gain
5
Companies
// Open to founder partnerships · worldwide

Got a bold AI idea?
Let's build it together.

I'm looking for founders building the next wave of AI-native products. You bring the vision and market insight — I bring the deep technical execution to ship a scalable production system from zero.

01
AI Product Development
LLM pipelines, RAG systems, AI agents, scoring engines — production-ready on FastAPI & AWS/GCP.
02
Scalable Backend Architecture
APIs and microservices built for 20K+ req/min. Distributed systems that grow with your product.
03
0-to-1 Execution
Move fast from idea to working MVP. I've shipped production systems solo and led high-output eng teams.
04
Remote & Global
Async-first, built for distributed teams. Open to equity, contract, advisory, or co-founder terms.
Domains I'm excited to build in
HealthTech & MedAI FinTech & InsurTech EdTech & Learning AI E-commerce Automation HR & Talent AI Data & Analytics Platforms Any AI-first SaaS
Let's Talk — I'm Ready to Build Call Me Directly →
Core Expertise

Technical Skills

Hover any skill to see where I've used it
Languages
Languages & Query
Python SQL
Backend
Frameworks & APIs
FastAPI Flask Django GraphQL REST JWT/OAuth2
Cloud & DevOps
Cloud Platforms
EC2 S3 Lambda RDS Redshift AWS Glue ECS GCP Docker
AI / ML
AI & Data Engineering
LangChain LangGraph RAG Pipelines Vector DBs (Pinecone/FAISS) Kong AI Gateway AWS Bedrock Azure OpenAI OpenTelemetry MCP Hugging Face CrewAI Claude / Anthropic API Gemini T5 LLMs PySpark NLP WhatsApp API
Databases
Database Systems
PostgreSQL MongoDB Parquet/S3
CI/CD
Dev & Deployment
GitHub Actions Jenkins Docker Git Microservices
What I Build

AI Systems: Agents, RAG & Fine-Tuning

Beyond infrastructure — I architect and ship AI-native products that people actually use, from stateful agentic workflows with LangGraph to fine-tuned domain models and RAG pipelines.

Applied ML · Shipped at Data Works
Fine-Tuned LLMs for Domain-Specific Scoring
Fine-tuned a T5-based model on labeled resume/job-description pairs to power a semantic candidate-scoring engine — outperforming generic embedding similarity on hiring-specific signal and lifting match accuracy by 40%.
T5 Hugging Face PyTorch Weighted NLP FastAPI Serving
Agent Orchestration
Stateful Multi-Step Reasoning
LangGraph state machines for multi-agent workflows with tool-calling, retries, and human-in-the-loop checkpoints — wired to external tools and data sources via MCP.
LangGraph LangChain MCP
RAG Infrastructure
Semantic Knowledge Systems
RAG pipelines on top of company knowledge bases and support docs — turning static information into a queryable, reasoning layer a product can act on.
Vector DBs Pinecone / FAISS RAG
Agent Platform
Domain-Specific AI Agents
Vertical agents built for specific workflows — sales follow-up, hiring screeners, patient intake — each with domain-tuned RAG and persistent user context.
CrewAI Persistent Memory FastAPI
Career Journey

Work Experience

Jul 2026 — Present
Staff Software Engineer
Everest Tech
  • Building an enterprise multi-provider AI Gateway on Kong Konnect (AI Gateway 2.x, hybrid self-hosted data plane) — a single OpenAI-compatible entry point for AWS Bedrock (Claude), Google Gemini (Vertex AI) and Azure OpenAI (GPT-4o).
  • Integrated MCP (Model Context Protocol) through the gateway so agents can securely call external tools and data sources, with the same auth, rate limiting and observability as model traffic.
  • Implemented gateway-level AI governance: prompt guard & semantic prompt guard, request/response transformers for PII / card-data masking, token rate limiting and per-client cost budgets.
  • Added semantic caching, RAG context injection backed by a Redis vector store, and OpenTelemetry tracing for end-to-end LLM observability.
  • Secured model access with Kong-issued API keys, OIDC and multi-issuer auth (Kong Identity + Okta); centralised all provider credentials in a vault so keys never leave the server side.
  • Designing a multi-agent system — an orchestrator that routes queries to a dedicated data agent or general agent — running entirely through the governed gateway.
Feb 2024 — Jun 2026
Staff Software Engineer
Data Works Technologies
  • Architected production-grade backend systems with FastAPI & Flask, integrating Hugging Face Transformers into scalable microservices deployed on AWS & GCP.
  • Achieved 99.99% uptime under heavy workloads across EC2, S3, Lambda, and ECS with auto-scaling strategies.
  • Built an AI candidate scoring engine using semantic similarity and weighted NLP algorithms, improving match accuracy by 40%.
  • Integrated CrewAI, Gemini, and T5-based LLMs for document understanding, resume parsing, and intelligent data extraction workflows.
  • Designed and deployed RAG (Retrieval-Augmented Generation) pipelines over resume, job-description, and internal knowledge-base corpora — using vector embeddings for semantic retrieval to ground LLM outputs and cut hallucinated responses in candidate/document Q&A workflows.
  • Developed ETL pipelines for millions of records using PySpark and AWS S3 Parquet, powering downstream Power BI dashboards.
Sep 2021 — Feb 2024
Senior Software Engineer
Mirafra Technologies
  • Built high-performance systems with FastAPI, Django, GraphQL for mission-critical products — TCRB, AARI, CureSoftware.
  • Architected APIs handling 2K+ requests/second with fault tolerance and horizontal scalability across microservices.
  • Designed real-time monitoring dashboards that reduced production debugging time by 45%.
  • Implemented end-to-end CI/CD pipelines via GitHub Actions and Docker, standardising deployment across dev, QA, and production.
  • Mentored junior developers in Python backend patterns, ORM optimisation, and containerised deployment best practices.
Jun 2019 — Jul 2020
Software Engineer
Walmart Labs
  • Engineered Flask-based microservices for supply chain workflows, improving backend processing efficiency by 25%.
  • Delivered 99.99% uptime handling 20K+ concurrent requests per minute across distributed clusters.
  • Reduced REST API response latency by 30% across internal and external integrations.
  • Implemented CD pipelines with Jenkins & Docker, compressing deployment time from hours to minutes.
Jun 2017 — Jun 2019
Software Engineer
IamHere Software Labs
  • Developed end-to-end backend systems with Flask and Django for data-driven web applications and internal tools.
  • Engineered an SMTP-based automated email delivery system — cut manual intervention by 70%.
  • Built custom reporting dashboards in Python + PostgreSQL enabling stakeholder data insights.
Key Achievements

Impact Highlights

01
AI Candidate Scoring Engine
Semantic similarity + weighted NLP algorithms to match candidates against job requirements at scale.
↑ 40% Match Accuracy
02
Real-Time Monitoring Dashboard
Django + PostgreSQL observability platform dramatically cutting production debugging cycles.
↓ 45% Debug Time
03
High-Throughput API Systems
Distributed API infrastructure handling 20K+ concurrent req/min with guaranteed 99.99% uptime at Walmart.
20K+ req/min
04
CI/CD Pipeline Automation
Jenkins & Docker pipelines that compressed multi-hour deployment cycles into fast, automated releases.
Hours → Minutes
05
LLM Document Intelligence
CrewAI, Gemini, T5-based LLMs for intelligent document understanding and resume parsing at enterprise scale.
Millions of Records
06
Email Delivery Automation
SMTP-based automated delivery system eliminating manual communication workflows at IamHere Labs.
↓ 70% Manual Work
Education

Academic Background

Master of Computer Applications
Bangalore University · 2013–2016
English · Hindi · Marathi · Kannada
Get In Touch

Let's build
something great.

Open to full-time roles, contract work, and founder partnerships in backend engineering and AI systems. Compensation is always open for discussion for the right role — if what you're building is ambitious, I want to hear about it.

📞 Call Me Now → Send a Message
+91 8880 144 666
Call directly — fastest way to reach me
hiremeforse@gmail.com
Email — usually reply within a day
linkedin.com/in/punitstaffengineer
LinkedIn Profile
Bangalore, India
Open to remote & hybrid roles