NishalChandra Reddy
For five years I've owned production systems end to end — architecting Java and Spring Boot microservices, Kafka event pipelines and LLM extraction that handle real traffic every day.
About
Software Engineer with 5 years building and owning production systems end to end — Java and Spring Boot microservices, Kafka event pipelines and LLM extraction pipelines running on live traffic.
I've led multi-service initiatives spanning six services and mentored engineers through architecture reviews. I'm comfortable at every layer — backend architecture, cloud infrastructure and IaC and React front ends.
M.S. in Computer Science from Arizona State University (2023–2025), with a track record of delivering reliable, high-performance software at scale.
Experience
Software Developer @ Ecohome.one
- Led the GDPR erasure program across six services — designed the specs and built a coordinator that fans a single deletion request out to idempotent erasure workers, backed by a MySQL state machine with per-household locks and single-active-request enforcement.
- Build and operate the AI receipt-extraction platform, a Node.js service on ECS Fargate that turns unstructured purchase emails into structured inventory records through a Gemini pipeline with vision OCR for PDF and image attachments.
- Hardened the pipeline against LLM failure with a four-tier fallback chain — per-stage heuristic extractors, a config-swappable Bedrock/Claude path and SQS re-queue with auto-requeue on provider recovery.
- Shipped two production services solo: an Alexa integration service (request-signature verification, OAuth account linking, circuit breakers, its own Terraform IaC and CI/CD) and EcohomeWeb, a React 18 + Vite SPA. Also built the eBay resale marketplace in the core Express API — listing state machine, scheduled delist workers and HMAC webhook verification.
Senior Software Developer @ Comviva
- Drove the architectural redesign of the MobiLytix Loyalty & Rewards platform — decomposing a monolith into modular Spring Boot services along Domain-Driven Design bounded contexts and rebuilding the campaign console as a React SPA. Lifted throughput 60% and feature delivery 35%.
- Built the service layer on Spring Data JPA & Hibernate over PostgreSQL, with Spring Cloud for centralized config and service discovery, Actuator metrics scraped by Prometheus and Spring Security with Keycloak (OAuth 2.0 / JWT) enforcing role-based access across tenants.
- Architected event-driven Kafka pipelines for asynchronous inter-service communication, reliably processing 700K+ daily transactions at sub-200ms latency.
- Migrated the platform from on-prem to AWS SaaS — containerizing with Docker and Kubernetes and adding centralized ELK logging, cutting infrastructure cost 40%.
- Led a cross-functional team of five, running architecture reviews and mentoring and standardized Maven multi-module builds with JUnit/Mockito coverage gates in CI.
Software Developer @ Comviva
- Built the Toll-Free Data Access system — network traffic monitoring in C/C++, Redis-backed session management and reverse-proxy routing in Node.js, serving zero-rated data access at carrier scale.
- Shipped rule evaluation, eligibility logic and reward fulfillment into the Java + Spring Boot campaign rules engine driving personalized reward programs.
- Automated CI/CD with GitLab CI, Jenkins and SonarQube, adding API contract tests and BDD — raising deployment frequency 40% and reducing production errors 30%.
- Instrumented services with Prometheus and Grafana for real-time oversight.
Projects
Swarm Intelligence Distributed Database
2025A scalable, fault-tolerant distributed database ingesting 10M+ IoT data points daily via Kafka and MongoDB sharding, with AI-driven query optimization using Ant Colony & Particle Swarm Optimization.
NetGuardian AI
2024End-to-end ML pipeline predicting severity of problematic internet use from 10,000+ hours of actigraphy time-series, using autoencoders and ensemble learning tuned with Quadratic Weighted Kappa.
Serverless Face Recognition
2024An end-to-end serverless pipeline triggered by S3 uploads with Dockerized ffmpeg for frame extraction, deploying ResNet-34 via OpenCV for real-time inference on 1GB+ videos without managing servers.
Tech Stack
Languages
Backend & APIs
Frontend
AI / ML
Data & Messaging
Cloud & DevOps
Testing & Auth
Observability
Let's build something.
Open to backend, distributed-systems and AI/LLM engineering roles. The fastest way to reach me is email.