GAN-Inspired RAG for the Legal Domain
Exploring how adversarial training principles can enhance retrieval-augmented generation systems for legal document analysis and case research.
Deep dives into distributed systems, AI engineering, and lessons from building real-world software.
Exploring how adversarial training principles can enhance retrieval-augmented generation systems for legal document analysis and case research.
Lessons learned from designing ShopMate — a scalable e-commerce system using event-driven microservices, NATS messaging, and Kubernetes.
A practical guide to multi-agent orchestration using LangChain and CrewAI, covering task delegation, memory sharing, and failure recovery.
How we built Shop Assist's real-time communication layer — integrating LiveKit for video and WebSockets for low-latency chat in an e-commerce context.
Simulating heat diffusion across a 3D grid using MPI and OpenMP, achieving significant speedup through domain decomposition and shared-memory parallelism.
Key architectural decisions behind Draftables — from authoritative server patterns to lag compensation and state synchronisation across clients.