Open to new roles · Kochi / remote

Harigovind
Valsakumar

Full-stack engineer, seven years in. I build the unglamorous middle — .NET Core services, Angular front-ends, Postgres schemas — and keep them running on Kubernetes.

Healthcare inventory, fleet telemetry, shipping logistics. Lately: predictive models in Python and the pipelines that ship them.

See the work Email me

Role: Full-stack / platform engineer
Since: 2019 · Gadgeon Systems Inc.
Core: .NET Core · Angular · PostgreSQL
Also: Python · Kubernetes · AWS

Years shipping production software

15+

Projects delivered, client and personal

Domains: healthcare, fleet, logistics

01 — About

Reliable systems, one layer at a time.

Seven years at Gadgeon Systems Inc. building multi-module platforms for healthcare, fleet and shipping clients — the kind of work where a bad migration is somebody's missing inventory at 2am.

Most of that has gone into microservices, third-party integrations and the CI/CD that makes releases boring. I care about the parts nobody demos: query plans, rollout strategy, the error path. Outside work I train models and build small tools.

2023 — Now

Platform, workflow & ML

BPMN workflow orchestration on Spring Boot and Flowable, Keycloak identity, and SageMaker-served anomaly models.

2021 — 2023

Fleet management & IoT

Extended a microservice platform to new device protocols and GPS telemetry ingestion; Helm-deployed .NET Core APIs with zero-downtime rollouts.

2019 — 2021

Healthcare & shipping systems

Angular and .NET Core inventory and logistics modules on PostgreSQL, synchronising data across distributed services and carrier APIs.

02 — Selected work

Three that mattered.

Client work at Gadgeon — details generalised where the contract requires it.

01Healthcare

Healthcare Inventory & Shipping System

A multi-module Angular and .NET Core application on PostgreSQL. Real-time synchronisation across distributed services, with logistics-provider integrations over REST.

Angular
.NET Core
PostgreSQL
REST APIs

ANGULAR · .NET · POSTGRES

02Fleet / IoT

Fleet Management Platform

Extended a microservices platform to support new IoT device protocols and GPS telemetry ingestion. .NET Core APIs on Kubernetes via Helm, tuned through CI/CD for zero-downtime rollouts.

.NET Core
Kubernetes
Helm
CI/CD

K8S · HELM · TELEMETRY

03Platform

Workflow Automation Platform

Full-stack orchestration built on React and Java Spring Boot, with Flowable as the BPMN engine, PostgreSQL for persistence and Keycloak for OAuth2/OIDC identity.

React
Spring Boot
Flowable
Keycloak

BPMN · SPRING · KEYCLOAK

Also built

ML / AI Superhero Character Predictor A CNN trained with Keras transfer learning on a custom-labelled image set, with augmentation for generalisation. ML / AI Credit Card Fraud Detection Binary classification on heavily imbalanced tabular data — SMOTE for balancing, evaluated on precision-recall AUC. ML / AI Linear Regression from Scratch Gradient descent, cost minimisation and evaluation written in NumPy, with no framework abstractions in the way. ML / AI Clustering Algorithms K-Means, DBSCAN and Agglomerative benchmarked across synthetic and real datasets with visualised boundaries. ML / AI Algorithms Library Classic CS algorithms in Python — sorting, graph traversal, DP, search — each annotated with its complexity. Tool JioFi Status Checker A native macOS menu-bar app in Swift that polls a router status API and surfaces battery and signal in the tray.

03 — Stack

What I reach for.

Front-end

Angular
React
TypeScript
HTML/CSS

Back-end & data

.NET Core
C#
Spring Boot
Python
PostgreSQL
SQL Server

Cloud & ML

AWS
Kubernetes
Docker
CI/CD
SageMaker
scikit-learn

04 — Contact

Let's build something dull and dependable.

Open to full-stack and platform roles. Also happy to talk shop about Kubernetes rollouts or model deployment.

contact@harigovindvalsakumar.com

GitHub LinkedIn X Instagram

HarigovindValsakumar