Sagar Hegde

Full-stack software engineer building production systems and AI agents across Python, TypeScript, and cloud-native infrastructure.

Worcester, MA·Open to relocation·F-1 OPT

Work

TotalPR

Autonomous code review agent for GitHub PRs: a 6-node LangGraph pipeline (FastAPI + Claude Haiku) that fetches diffs via GitHub App auth, runs sandboxed static analysis in isolated Docker containers, retrieves labeled feedback from a pgvector memory store, and posts structured findings back to the PR. Labels close the loop: rejecting a finding suppresses it on the next run of the same diff.

F1 0.86 (precision 0.82, recall 0.90) · ~$0.005/review · p95 11.3s

FastAPI · LangGraph · Claude Haiku · pgvector · Docker · Fly.io · OpenTelemetry

GitHub→

Mail Lens

Phishing email analyzer combining a six-check deterministic rule engine (SPF/DKIM/DMARC, lookalike domains, URLs, attachments) with an LLM assessment to produce a single 0–100 explainable risk score. Degrades to rules-only when the model API is unavailable; validated under chaos tests.

92% precision on 500-email labeled set · p95 1.8s end-to-end

TypeScript · React · FastAPI · Python · OpenAI API · GitHub Actions

Live demo→

System Design

Cellular Data Monitoring System

WPI capstone with an industry sponsor: architected a cellular telemetry system for 12 HVAC sites, designed to scale to 50+. Produced an 11-entity domain model and a two-tier Azure architecture spec from a multi-week requirements engagement with the sponsor CTO.

Negotiated 5 proposed traffic categories down to 3, backed by production data

Systems design · Azure · Requirements engineering

Read report→

Experience
CoolGreenPower · Software Engineer

Internal React admin tooling and Python pipelines processing 1.4M sensor events/day across 12 HVAC sites.

May 2025 – Apr 2026
IFTAS · Cloud Services Engineer

Multi-tenant Kubernetes operations and platform SOPs for a financial-infrastructure provider.

Aug 2023 – Jun 2024
Swegon BlueBox · Software Engineer, Web

Vue real-time telemetry dashboards and alarm-configuration tooling for HVAC field engineers.

Sep 2022 – Aug 2023
About

I'm a software engineer who enjoys building backend systems, full-stack applications, and AI-powered tools. Over the past few years I've worked across Python, TypeScript, React, Azure, and cloud infrastructure, with a growing focus on production AI systems. I like working on the engineering around LLMs just as much as the models themselves, evaluation, observability, reliability, and deployment. I'm looking for opportunities to build software that's practical, maintainable, and used by real people.