Hello, World! 👋

Ankitha Suresh

Software Engineer · Backend & AI/ML · Cloud & Infra

MS CS @ UMass Amherst · Ex HPE · Building scalable systems & AI-powered products

About Me

Who Am I?

I'm a Software Engineer with a Master's in Computer Science from the University of Massachusetts Amherst (GPA: 3.93 / 4.0). I specialize in building scalable backend systems, AI/ML pipelines, and cloud infrastructure.

With experience at Hewlett Packard Enterprise and Headstarter, I've shipped production systems that cut cluster recovery time by 98%, built RAG pipelines serving 500+ daily users, and engineered event-driven architectures at 99.5% uptime.

Outside of work, I love solving DSA problems on LeetCode, building side projects that solve real problems, and exploring new corners of the tech world.

M.S. Computer Science

University of Massachusetts, Amherst

GPA 3.93 / 4.0 · May 2025

B.E. Computer Science

JSS Science and Technology University, India

GPA 9.34 / 10 · May 2021

My Skills

What I Work With

Programming

Python Java TypeScript JavaScript SQL Bash

Frontend

React Next.js HTML / CSS

Backend & APIs

Node.js FastAPI Flask gRPC GraphQL REST

AI / ML

LangChain OpenAI API Pinecone RAG Vector Databases

Databases

PostgreSQL MongoDB MySQL Redis

Cloud & Infra

AWS Docker Kubernetes Terraform GitHub Actions GitLab CI/CD

Testing & Tooling

Pytest JUnit Postman Apache Kafka CloudWatch

Work Experience

Career Journey

Headstarter

Software Engineer Resident (After Fellowship)

June 2024 – Dec 2025

Architected a multi-stage RAG pipeline (LangChain + Pinecone) cutting query failure rate from 18% → 4% across a 50K-document knowledge base serving 500+ daily users.
Built an AI code review assistant (React, Node.js, OpenAI) processing 1,000+ PRs with a 78% developer acceptance rate on suggested changes.
Optimized RAG query latency from 3.2 s → 800 ms p95 via async retrieval, result caching, and request batching at 99.5% uptime.
Built a multi-step LLM agent using OpenAI tool-calling that reduced manual triage time by 60%.

PythonNode.jsReactLangChainPineconeOpenAI

Hewlett Packard Enterprise

Software Engineer

Sept 2021 – July 2023

Engineered event-driven failure detection service cutting cluster recovery time by 98% (12 min → 14 s), directly reducing customer-facing downtime.
Built unified CLI tool for cluster operators replacing 4 ad-hoc scripts and cutting average operator task time by 35%.
Developed cluster monitoring dashboard across 200+ clusters, reducing time to identify degraded nodes from 45 min → under 2 min.
Rewrote disaster-recovery validation suite with concurrent execution across 50+ clusters, cutting runtime from 5 h → 55 min.

JavaPythonREST APIsDockerKubernetesAnsibleTerraform

Hewlett Packard Enterprise

R&D Engineer Intern

Feb 2021 – Aug 2021

Built Python + Ansible installation pipeline that ran prerequisite checks, auto-remediated failing conditions, and reduced deployment time from 1–2 h → under 15 min.
Diagnosed and fixed a concurrency bug silently inflating MTTR across 50+ clusters, reducing incident recovery from 25 min → 10 min.
Contributed health-check and configuration validation modules replacing 4 h of manual cluster setup with a 25-min automated run.

PythonAnsiblePrometheusGrafana

My Work

Featured Projects

Kreede

Smart Court Booking System

Full-stack court booking and customer management platform for a live pickleball business. Real-time slot scheduling for 200+ users, eliminating manual booking entirely.

ReactNode.jsPostgreSQLREST API

Algocards

AI-Powered DSA Flashcards

AI-powered platform for mastering Data Structures & Algorithms. Features curated flashcard sets (Basic DSA, Algorithms, Advanced DSA) and interactive assessments.

Next.jsReactOpenAITypeScript

US Election Sentiment Analysis

Real-Time NLP Pipeline

Real-time sentiment analysis pipeline ingesting and classifying 10,000+ live social media posts on the 2024 US election, surfacing results on a live dashboard.

PythonNLPStreamingDashboard

Placement Management System

University Placement Portal

Placement platform used by 400+ students at JSS University. Role-based access, application tracking, and recruiter dashboards actively used across placement season.

Full-StackRole-Based AuthPostgreSQL

ZenMode

Mood-Adaptive Video Feed

Instagram-like app with a mood-based video feed (MoodFeed). Users select their current mood and receive a dynamically personalized video experience aligned to their emotional state.

Next.jsReactAITypeScript

Car Make & Model Prediction

Computer Vision / Deep Learning

CNN-based image recognition model using Inception-v3 to predict car make and model from images. Achieved 92% accuracy on a large automotive image dataset.

PythonCNNInception-v3TensorFlow

Contact

Get In Touch

Have an opportunity or just want to say hi?
My inbox is always open.

s.ankitha0904@gmail.com

+91 761 968 0294