SANTHOOPA [01] JAYAWARDHANA.
Software engineer and AI practitioner working at the intersection of scalable product engineering and intelligent systems — shipping agentic AI, distributed data pipelines, and full-stack platforms for global teams.
Scaffolding the boring parts. Wiring up the smart ones.
I'm a software engineer and AI practitioner who thrives at the intersection of scalable product engineering and intelligent systems. Over 5+ years I've designed and delivered production-grade solutions across full-stack platforms, real-time data pipelines and AI-powered ecosystems — from global coworking platforms and telecom automation to enterprise sales tools and voice-enabled multi-agent AI.
I own delivery end-to-end — from React, Next.js & Angular frontends to Node.js (NestJS) & Java backends, with GraphQL, Apache Kafka and AWS underpinning the architecture. I design and integrate agentic AI using LLMs, RAG with Pinecone and tool-based retrieval workflows built on LangGraph, and I've led complex third-party integrations across payments, CRM and communications platforms to power real business workflows.
Beyond industry work I'm a published researcher with peer-reviewed contributions in deep learning and NLP — outbreak-related tweet detection for public-health surveillance, hate-speech detection in Sinhala (a low-resource language), and real-time traffic-flow estimation with YOLOv4 + SORT, presented at ISCRAM 2021 (Virginia Tech). Academically, I hold a First Class Honours BSc (Hons) in Management & IT — awarded the Gold Medal for best academic performance — and a Postgraduate Diploma in Data Science & AI.
DOSSIER OPEN
STACK ◆ ACTIVE
Systems shipped. Mostly in production.
Where I've put hours in.
Trained in theory. Built for practice.
BSc (Hons) in Management & Information Technology
Postgraduate Diploma, Data Science & Artificial Intelligence
Papers & publications.
Thinking out loud.
Spec-Driven Development (SDD) with Coding Agents
Why vibe coding fails for real projects — and how writing specs first keeps coding agents building what you actually meant.
DATA SCIENCE COLLECTIVE · JUL 2026 · READ ↗ DEEP LEARNING · NLPAI to Detect Hate Speech related Sinhala Comments in Social Media
Using deep learning models to detect Sinhala hate speech on social media — applying NLP to a low-resource language.
TOWARDS AI · NOV 2020 · READ ↗ DEEP LEARNING · SEQUENCE MODELSSequence Models & Recurrent Neural Networks (RNNs)
Understanding deep recurrent neural networks — how sequence models work and where RNNs fit.
TOWARDS DATA SCIENCE · JUL 2020 · READ ↗Tools of the trade.
Languages + Frameworks12
Data + Infra08
AI / ML12
Integrations + Platforms08
What I can build for you.
Start a project ▸
Full-Stack Product Engineering
Ship production web apps end to end — from polished, responsive frontends to scalable backend services and APIs.
Agentic AI & LLM Systems
Design and integrate LLM-powered agents with retrieval, tools, and memory — from prototype to reliable production workflows.
Real-time Data Pipelines
Event-driven streaming and data flows for CRM, billing, and analytics that stay reliable at scale.
Third-party Integrations
Wire up payments, CRM, and comms into seamless user and business workflows with robust error handling.
Cloud & Technical Leadership
Cloud-native architecture, clean code, and end-to-end delivery — leading complex builds with global stakeholders.
ML & Applied Research
NLP, computer vision, and deep learning — from research prototype to deployed, measurable outcomes.
Side experiments, built for fun.
AI Lab — 7 live experiments
Gesture control, object detection, face reading, cartoonization, pose tracking, a live puppet face and speech recognition — small models running entirely on your device. No servers, no tokens.
7 EXPERIMENTS · $0.00/REQUEST · OPEN ↗ GAME · WEBGLdebris.field
A short arcade break between sections — dodge, shoot, and survive as long as you can.
1 LEVEL · NO INSTALL · PLAY ↗