Sahan Serasinghe

Senior Software Engineer | Master of Data Science

Projects

Selected work across developer experience, cloud, and systems.

Side projects, open source, and academic buildsβ€”kept concise, with links to code, case studies, and demos.

πŸ§‘β€πŸ’» Notable OSS

Personal builds and open source contributions with community usage.

πŸ§‘β€πŸ’» Open Source

In-Memory Caching .NET with IMemoryCache (39 ⭐️)

Improving performance of monolithic .NET applications with in-memory caching.

.NETC#ASP.NET CoreIMemoryCache

πŸ§‘β€πŸ’» Open Source

GitHub API for Java (1K ⭐️) β€’ Contributor

I contributed to the GitHub API for Java library by adding support for allowing Fork type to be specified when performing a search.

JavaGitHub APIGitHub

πŸ§‘β€πŸ’» Open Source

Distributed Caching in ASP.NET Core with Redis (23 ⭐️)

Improving performance of .NET microservices applications with distributed caching.

.NETC#ASP.NET CoreIDistributedCacheRedis

πŸ§‘β€πŸ’» Open Source

Understanding WebSockets with .NET (13 ⭐️)

I presented a reference implementation to understand RFC 6455 WebSockets specification and configure a generic ASP.NET (Core) 5 application to communicate over WebSockets connection with SignalR.

.NETC#ASP.NET CoreWebSocketsSignalR

πŸ§‘β€πŸ’» Open Source

gRPC on Go (12 ⭐️)

A simple gRPC web service written in Go without using third-party libraries.

GogRPCProtocol Buffers

πŸ§‘β€πŸ’» Open Source

gRPC on .NET (6 ⭐️)

Code samples for the Introduction to gRPC series in .NET.

.NETC#gRPC

πŸ§‘β€πŸ’» Open Source

Adding Intelligence to React Apps with Azure Cognitive Services

Enhance your React apps with AI-powered features.

ReactAzure Cognitive ServicesNext.js

πŸ§‘β€πŸ’» Open Source

Go Azure Functions (8 ⭐️)

Playground for deploying Azure Functions written in Go.

GoAzureServerless

πŸ§‘β€πŸ’» Open Source

Kafka Docker (8 ⭐️)

Quick way to spin up a local Kafka instance with Docker.

KafkaDockerStreaming

⚑️ Side Projects

sahansera.dev

This blog! Over 300K pageviews (as of March 2023), 36 posts with over 14K+ monthly active users. Well received in the .NET & Go Communities.

ReactGatsbyPWAServiceWorkerAnt DesignTypeScriptGraphQL

⚑️ Side Projects

Medline Pubmed Data Extractor

A simple .NET console application that will process a given Medline Pubmed dataset (XML) and generate TSV files with given attributes.

.NETC#LINQTSVXML

πŸŽ“ Academic Projects

Notable work from my master's and undergrad yearsβ€”visualization, ML, and hardware.

πŸŽ“ Academic Projects

Narrative Visualization with D3.js

One of my notable achievements was the creation of a captivating narrative visualization, entirely powered by D3.js and vanilla ES6, without relying on any packaging tool. I presented a single-page application architecture, enabling seamless navigation using both keyboard and pointing devices, akin to a presentation tool. To ensure data integrity and efficiency, I used Python's Pandas library for dataset preprocessing.

D3.jsJavaScriptHTMLCSSPythonPandas

πŸŽ“ Academic Projects

Twitter Sarcastic Tweet Detection using Deep Learning

During my Master's degree, I had the opportunity to lead a talented team in the highly competitive Text Classification Competition for Sarcasm Detection in Tweets. Our team scored an impressive F1 score of 0.754. In this competition, my primary focus was on the crucial tasks of dataset preprocessing and training a powerful BERT model.

PythonPyTorchDeep LearningNatural Language ProcessingBERTTextBlobTransformers

πŸŽ“ Academic Projects

Self-driving car using Raspberry Pi

I built a self-driving 4WD car leveraging the SunFounder's PiCar kit. Throughout this project, I lead the implementation of the mapping module using the onboard ultrasonic sensor readings. As the project progressed, I further expanded its capabilities by incorporating features like obstacle detection (sign posts, walls, etc.), route mapping, and orientation detection.

Raspberry PiPythonComputer VisionOpenCVUltrasonic SensorMatplotlib

πŸŽ“ Academic Projects

LogXTractor

My final year undergraduate research project where I researched, designed and implemented a web based IDE for extracting actionable insights from log files. I presented a fully working prototype working with large scale log (>2GB) files and inferencing patterns using ML techniques. This project also won the best academic research project award for 2014 and best academic paper award at the ICIIT 2016 conference

CPythonMachine LearningLog AnalysisJavascriptHTMLCSS

πŸŽ“ Academic Projects

MobileHub Question Answering System

Operating as a full stack application, it has a client-side application implemented using vanilla JavaScript, a backend REST API powered by PHP, and a MySQL database for seamless data management. Features implemented to ask and answer questions, engage in voting, utilize tags, close questions, manage users, access an admin panel, and even employ voice search functionality. From inception to completion, I designed and developed every component of this project, leading it to secure the highest marks within my batch.

PHPMySQLHTMLCSSREST APIJavaScriptjQueryBootstrap