Software Engineer  ·  Open to Relocation

Antarang
Poogalia

Backend core and full-stack reach — building software that removes friction from everyday decisions.

 Open to Work
H1B Visa (Transferable)
Resume
Scroll

About

Built on curiosity.
Grounded in engineering.

Antarang Poogalia

I got into software engineering because I wanted to solve real problems. The first thing I ever built was a website for my father to present his products to international customers. That instinct to find the right tool and build something that actually helps has driven everything since.

Today I work at ArcelorMittal Calvert building the systems that keep a major steel plant running, from real-time industrial hardware integrations to ML-driven warehouse optimization. Before that I built data pipelines at Hearst and modernized pharma software at ZS Associates.

🔨
Backend core: systems integrations, database optimization, cloud pipelines, C#/.NET and Java at production scale.
🌐
Hands-on across frontend, AI and ML, and cloud. I pick up what the stack needs and go deep fast.
🤖
Currently exploring how AI agents, MCP, and NLP can power smarter recommendation systems that remove real friction from everyday decisions.
📄
Published researcher. Built an NLP spam classifier in 2020 before LLMs were mainstream, achieving 96% accuracy on 11K+ messages. Published in BBRC.
Currently Learning
Building agentic recommendation systems with LangGraph and diving into Model Context Protocol to understand how future AI will interact with the world.
What I Want to Work On
Software that removes friction from everyday decisions. Trip planning, personal logistics, things people dread figuring out. The kind of product where the best version is one users barely notice because it just works.

Experience

Where I have worked.

A track record of shipping impactful software across manufacturing, media, and pharma.

Sept 2024 to Present
ArcelorMittal Calvert
Calvert, AL
Software Developer
  • Enhanced the Warehouse Management System with visual tooling for 100+ daily operators.
  • Built a RAG-powered chatbot that reduced recurring support questions and accelerated onboarding.
  • Boosted database read/write performance by ~20% through C#/.NET indexing improvements.
  • Engineered real-time PLC messaging between coil car machinery and the management system, enabling live start, stop, and reset control with zero safety incidents.
  • Deployed ML algorithms with international teams, improving yard storage efficiency by ~40%.
  • Automated deployment pipelines, reducing release downtime by ~30%.
A problem worth calling out
Real-time control of industrial hardware at a steel plant
One of the more unusual problems I have solved: building real-time messaging between a coil car — a machine that physically moves heavy steel coils between factory bays — and our warehouse management system. Getting start, stop, and reset signals from industrial PLC hardware to talk reliably to software in a live production environment, where a missed message has physical consequences, required a different kind of careful thinking than typical web engineering.
Jun to Aug 2023
Hearst Communications
Remote
SDE Intern
  • Modernized Concur expense retrieval for 20,000+ users through a unified HR interface.
  • Built S3 to SQL Server ingestion via AWS Lambda, saving ~1 hour of manual work daily.
  • Reduced expense retrieval response time by ~30% via SQL query optimization.
  • Integrated backend APIs with the UI, enabling HR tasks to be completed ~40% faster.
Mar 2021 to Jul 2022
ZS Associates
Pune, India
Software Engineer
  • Modernized pharma software using Angular, improving usability for ~10,000 users.
  • Designed ~250 test cases across functional, API, and ETL testing, uncovering ~100 pre-release defects.
  • Scaled Java-based UI automation framework, cutting manual QA effort by ~40%.
  • Contributed to CI/CD integration that doubled deployment speed.

Skills & Projects

What I know. What I have built.

Click any skill to spotlight matching projects. Click again to clear.

Languages
Frameworks
AI / ML
Cloud
📦
Inventory Manager
Cross-platform mobile app with barcode scanning, image upload, and real-time search.
React NativeNode.jsMongoDB
🆎
Restaurant Finder Agent
LLM-powered agentic system surfacing restaurants from nuanced preferences.
PythonLangChainML
💼
Job Recommender
Resume-to-job NLP matching engine surfacing the most relevant listings.
PythonNLPScikit-learn
🔍
Job Scraper
Aggregates recent job listings. React frontend, Java + Spring backend.
JavaSpringReact
🌎
SwachhMap
Civic tech platform for reporting waste issues with geolocation mapping.
ReactNode.jsMongoDB
📩
Spam Filter
Android + Python ML classifier. 96% accuracy on 11K+ messages. Published in BBRC.
AndroidJavaNLP📄 Published
🚗
ParKar — Smart Parking
Parking slot reservation with GPS verification and PayPal Sandbox gateway.
AndroidJavaFirebase
📱
Phone Recommender
ML pipeline ranking phones by user preference and filters.
PythonML
📊
Risk Viz Platform
D3.js analytics with stream graphs, circle packing, geospatial risk mapping.
D3.jsJavaScript
96%Accuracy
Published Research · Personalized Spam Filter for Social Networks
In 2020, before LLMs were mainstream, I built and published an NLP-based spam classifier trained on 11,000+ social network messages. Using Naive Bayes, Logistic Regression, and SVM, it achieved 96% accuracy on unbalanced data. Published in BBRC.

Contact

Buzz me.

Drop your LinkedIn or email and I will reach out. No contract roles please.

Enter your LinkedIn profile URL or email address
Please enter a valid LinkedIn URL or email address
Message (optional)
📳

Buzz sent! I will reach out soon.

9:41
▶ ■ ■■■
Recruiter
Hey Antarang, saw your portfolio!
Would love to connect 🙌
iMessage
▲