Bengaluru, India · Available for meaningful problems

I build dependable systems for complex ideas.

I'm Sumit, an AI Platform Software Engineer at Oracle. I work across distributed systems, Java and Python services, cloud reliability, automation, and the practical edge of AI.

Sumit Samantray
AI platforms
Cloud reliability

2+ years delivering AI platform systems

100+ regions supported in cloud operations

99.95% uptime for Oracle AI Language Service

Systems thinking, with a builder's instinct.

My work sits where software engineering, cloud infrastructure, and applied AI meet. I enjoy turning unclear, high-stakes problems into reliable systems with measurable outcomes.

At Oracle, I work on the platform behind AI language services: building Java and Python microservices, strengthening deployment and observability practices, and helping systems stay dependable at scale. Outside the platform, I build tools that make a focused workflow feel simpler—like PegShot.

Based in
Bengaluru, India
Education
IIT Madras · Dual Degree
Focus
AI platforms · distributed systems · reliability

From aerospace and data science to AI platforms.

A path shaped by curiosity, research, and building systems that operate when the stakes are real.

2019

IIT Madras

Dual Degree in Data Science and Aerospace Engineering

2022

EXL Service

Data Science internship · automation and data processing

2023

Oracle

Member of Technical Staff intern · NLP data curation

2023–24

AI Research

IIT Madras · deep learning and chess intelligence

2024—Now

Oracle AI

AI Platform Software Engineer · reliability at global scale

A practical stack for building, operating, and improving.

Tools matter most when they help a team ship with confidence. These are the technologies I reach for across platform work, reliability, and product engineering.

01

Backend & systems

Java, Python, C++, SQL, Spring Boot, Django, Kafka, REST, data structures, OOP, SOLID, distributed systems.

02

AI & data

LLMs, RAG, BERT, LangChain, prompt engineering, model evaluation, MongoDB, relational databases, data pipelines.

03

Cloud & reliability

OCI, AWS, GCP, Azure, Kubernetes, Docker, Helm, Terraform, Ansible, Prometheus, Grafana, ELK.

04

Delivery & product

CI/CD, Jenkins, GitHub, Git, TestRail, React, TypeScript, system design, root-cause analysis, Agile delivery.

Reliability is a feature, not an afterthought.

01

Helped support Oracle AI Language Service across 100+ cloud regions, 1,500+ OCI Linux instances, and 300+ load balancers while maintaining 99.95% uptime.

02

Expanded supported language capability by 10× through a multilingual LLM backend migration, targeted prompt engineering, and model evaluation.

03

Automated CVE remediation with graph dependency resolvers, saving approximately 20 hours of manual effort each month.

04

Co-led a rapid recovery of the OCI Language Service after a large-scale event affecting 100+ customers.

Chrome extension · product engineering

PegShot

Web capture should not force every idea into a rectangle. PegShot lets people select custom-shaped or scrolling webpage regions with four pegs, permanently redact sensitive details, extract text, copy a PNG, or export a PDF.

  • Manifest V3 Chrome extension
  • Local image composition and OCR workflow
  • Server-backed trial and license entitlement
Visit the PegShot landing page

Learning stays useful when it reaches production.

2024

OCI DevOps Professional with Containers and Kubernetes

Oracle
2024

Oracle Cloud Infrastructure Architect Associate

Oracle
2024

UNIX and Linux Essentials

Oracle
2019

JEE Advanced All India Rank 3008

Top 0.24% of candidates

Start a conversation

Have a difficult system or a useful idea? Let's build it well.