arc

Ayush Ravi Chandran

आयुष रवि चंद्रन  ·  ஆயுஷ் ரவி சந்திரன்

M.S. Computer Science @ UMass Amherst  ·  New Grad 2027

Currently: SWE Intern @CDS-AI

about

B.S. in CS + Math from UMass Amherst, now doing my M.S. in CS there as a Bay State Fellow. I build things that range from data pipelines serving researchers to poker-playing agents - I just like making stuff that works well.

Along the way: Chancellor's Scholarship, Dean's International Scholarship, Flynn Research Scholarship, Dean's List, and an Outstanding UCA Award for CS 240.

I care about backend systems that are correct every time, not just most of the time. ML and security live nearby; I like that both punish shallow understanding. Outside of code: games, anime, gym, and a tendency to go one tab too deep on things adjacent to whatever I'm building.

Looking for new grad software engineering roles starting 2027. .

Based in

Massachusetts, USA

School

UMass Amherst

Degree

M.S. CS '27 · B.S. CS & Math '26

Fellowship

Bay State Fellow

Focus

Backend · Data · ML

Open to

New Grad 2027 roles

Ayush
“

Programs must be written for people to read, and only incidentally for machines to execute.

Harold Abelson, SICP

work

Sep 2025 - Present

Software Engineering Intern

CDS-AI ↗

Replaced third-party tooling and manual workflows with a full-stack environmental monitoring platform serving 100+ researchers across Massachusetts DEP-funded salt marsh research.

  • Sole engineer on the Salt Marsh Data Platform (saltmarshdata.org), owning the web portal, data pipelines, and mobile app in production
  • Built backend data pipelines (PostgreSQL, Azure/S3 storage) to ingest, standardize, and store multi-organization salt marsh data for the Massachusetts DEP-funded Salt Marsh Data Platform, supporting 100+ researchers across 20+ organizations
  • Implemented fault-tolerant NAS synchronization across heterogeneous formats with schema mapping and incremental batch processing
  • Developed a multi-tenant web portal (React, FastAPI, PostgreSQL) with five-tier role-based access control and containerized deployment, managing 1,000+ weekly field records
  • Built an offline-first Flutter mobile app for field data collection replacing third-party tooling - local SQLite storage, GPS capture with RTK location matching, nested survey workflows, and automated QC validation before ingestion
  • Built containerized React tools with interactive geospatial visualizations for Woodwell Climate Research Center
ReactTypeScriptFastAPIFlutterPostgreSQLSQLiteAzureS3DockerGeospatial Data

Sep 2026 - Present

Supporting 200+ students in probability and statistics as part of the Bay State Fellowship.

  • Teaching Assistant for COMPSCI 240 (Reasoning Under Uncertainty), appointed through the Bay State Fellowship
  • Lead weekly discussion sections and hold office hours for 200+ students
  • Prepare exam questions and grade assignments and exams
ProbabilityStatisticsBayesian ReasoningTeaching

May 2025 - Aug 2025

Software Engineering Intern

MGHPCC (Unity HPC Cluster) ↗
repo ↗

Found $120K+ in wasted GPU resources and cut log analysis time by 80%, directly improving resource allocation across a major HPC cluster.

  • Engineered low-latency telemetry pipelines in a Linux HPC environment to process 12+ months of cluster activity
  • Optimized SQL queries and indexing strategies to accelerate SLURM log analysis by 80% across 10.9M+ job records
  • Applied DBSCAN clustering to identify 63,000+ underutilized A100 GPU hours, uncovering $120K+ in wasted compute
  • Built CI/CD pipelines with GitHub Actions and pytest for monitoring services
PythonSLURMSQLLinuxCI/CDGitHub ActionsDBSCANHPC

May 2025 - May 2026

Recognized with the Outstanding UCA Award while teaching probability and guiding teams through building production-grade distributed systems.

  • Won the Outstanding UCA Award for contributions to COMPSCI 240
  • Mentoring student teams 1-on-1 in CS426 (Scalable Web Systems) on distributed systems projects involving microservices, Redis, caching, and React
  • Held discussion sections and office hours for 150+ students in CS240, covering probability, statistics, and reasoning under uncertainty
  • Clarified complex topics including Bayesian reasoning, probabilistic modeling, and statistical inference through guided problem-solving
MicroservicesRedisCachingReactProbabilityBayesian ReasoningMentorship

Jun 2021 - May 2025

Head of Technical Operations

Instilt Educate ↗
repo ↗

Streamlined onboarding, improved developer velocity, and saved 10+ hours/week through automation and re-architected backend systems.

  • Led a team of 10+ engineers across weekly sprint meetings
  • Managed technical operations and system access control for 300+ users
  • Built automated onboarding pipelines and scheduling automation with Google Calendar and Zoom APIs, saving 10+ hours/week
  • Led backend re-architecture for scalable, high-performance systems
JavaScriptNode.jsAutomationCloudLeadership

projects

“

The purpose of computing is insight, not numbers.

Richard Hamming

research & academic work

Applied Research

Unity HPC Cluster Usage Analysis

CDS-AI Summer Research · Summer 2025

Built a Python analytics module for HPC workload telemetry and presented findings on usage patterns and resource allocation in the Unity HPC cluster. Sponsored by Adobe Research.

Paper

Exploiting CVE-2016-5764: Stack Buffer Overflow in Rumba FTP Client

UMass Amherst · Fall 2025

Analyzed a stack-based buffer overflow in the Rumba FTP client, documented the SEH overwrite exploitation path, and summarized the reverse engineering findings in a companion writeup.

Paper

Physics-Informed Neural Networks for Non-Equilibrium Stochastic Systems

UMass Amherst - Independent Study · Fall 2025

Studied stochastic systems using Fokker-Planck equations, comparing classical numerical solvers and physics-informed neural networks on double-well and non-equilibrium dynamics. Analyzed accuracy, stability, and scalability in high-dimensional diffusion problems.

contact

Let's connect.

Whether it's a job opportunity, collaboration, or just a question - I'm open to it.

Ayush Ravi Chandran · 2026Built with Next.js + Tailwind