People detection and optional tracking with Tensorflow backend.

A powerful and user-friendly web application that allows users to easily download audio content from Spotify URLs. This application is built using Flask for server management and spotDL for asynchronous audio file downloading.

Aether-SPARC v3: Asynchronous neuromorphic signal processor integrating Selective SSM (Mamba) with Predictive Coding. Achieves ~88% compute reduction by executing state updates only on signal surprise. Verified with hardware-accurate MAC accounting and Intel Loihi 2 silicon energy projections. High-sparsity edge DSP.

An asynchronous online judge built on a distributed producer–consumer architecture using BullMQ, Redis, Docker, and MongoDB.

Serverless image processor built with Go, AWS Lambda, S3, and SQS. Automatically creates thumbnails and medium-sized versions of uploaded images with optimized compression.

A robust, scalable payment gateway system built in Java, designed to demonstrate Low-Level Design (LLD), clean architecture, and real-world backend engineering concepts such as idempotency, asynchronous processing, and webhook reliability.

.NET 9 cloud backend for asynchronous podcast transcription with ASP.NET Core, FFmpeg, Azure AI Speech, Blob Storage, and Cosmos DB.

RideSphere is a scalable, event-driven ride-sharing backend system built using microservices architecture with Node.js, Express, MongoDB, RabbitMQ, and JWT authentication. The platform demonstrates real-world system design principles, asynchronous communication, API Gateway pattern, and fault-tolerant service orchestration.

API assíncrona e resiliente de processamento de filas desenvolvida com NestJS, BullMQ e Redis. Inclui tratamento de falhas com DLQ, concorrência, retentativas automáticas, documentação interativa com Scalar, testes com Jest e suporte a Docker.

A distributed J2EE Online Auction System utilizing EJB (Enterprise Java Beans) (Singleton, Stateless, MDB) & JMS for high-concurrency, asynchronous bid processing. Features realtime updates via WebSockets, thread-safe concurrency control, & a Bootstrap/jQuery UI powered by Jackson JSON serialization, deployed as a scalable EAR (Enterprise Archive)

Simplified API for seamless OpenAI requests, streamlining access to AI capabilities... Created at https://coslynx.com

Production architecture documentation for implementing Redis and Celery in Django applications, covering caching, asynchronous task processing, environment isolation, Kubernetes deployment, and engineering design decisions.

Kafka Laravel Queue – A Laravel package that integrates Apache Kafka as a queue driver, enabling efficient, scalable, and event-driven job processing in Laravel applications. 🚀

People detection and optional tracking with Tensorflow backend.

PulseWatch is a distributed uptime monitoring platform that schedules website health checks, processes them asynchronously through RabbitMQ workers, tracks latency and failures, manages incidents and recovery, and delivers real-time service status updates.

Docker Compose setup for a highly available Kafka cluster using KRaft (no ZooKeeper). Includes Kafdrop for web-based Kafka monitoring.

Java demo showcasing both the Observer and Publish/Subscribe design patterns. Includes simple, practical implementations to highlight their differences.

Leverage Azure AI Document Intelligence to extract text, tables, and key data from complex forms and automatically update your enterprise database.

Agibank challenge | Scalable financial contract validation service with high-volume asynchronous processing

An asynchronous fifo keeps two independent clocks: one for writing and one for reading.