Principal Software Engineer · Lahore, Pakistan 🇵🇰
Six-plus years building and scaling multi-tenant SaaS platforms in Ruby on Rails and React — and increasingly, putting LLMs into production where they earn their place.
I like the unglamorous parts of engineering: untangling legacy monoliths, cutting response times, making deploys boring, and getting AI features past the demo stage into something that holds up under real users and real edge cases.
Most of my recent work sits here — building on top of LLMs rather than training them, with the same scrutiny I'd apply to any production system.
Retrieval-augmented generation (RAG) — vector databases for semantic search and long-term context, so responses stay grounded in real data instead of hallucinated.
Conversational systems with memory — persistent, contextually aware conversation state across sessions rather than one-shot prompts.
Document intelligence — AI-powered extraction from unstructured documents, auto-populating structured forms and cutting manual data entry out of clinical workflows.
Multimodal / 3D interfaces — pairing LLM backends with real-time 3D avatars (React Three Fiber) for immersive, human-feeling interactions.
Orchestration and pipelines — LangChain for chaining, tool use, and structured outputs; prompt design, evaluation, and the guardrails that keep responses predictable.
Toolkit: OpenAI LangChain Vector DBs RAG Embeddings Prompt Engineering React Three Fiber
Backend · Ruby on Rails · Node.js · Python · GraphQL Frontend · React.js · Apollo Client · JavaScript · HTML/CSS Data · PostgreSQL · MySQL · Amazon RDS · DynamoDB · Redis Cloud & DevOps · AWS · Azure · Docker · Sidekiq · CI/CD Testing · RSpec · Apollo GraphQL
Scaling SaaS platforms. Multi-tenant architectures serving thousands of users — keeping response times sub-second as the user base grows.
Untangling legacy code. Refactoring monoliths into something maintainable, planning monolith-to-microservices migrations, and building the test coverage that makes those changes safe to ship.
Cloud migrations and cost work. Moving workloads between infrastructure models, right-sizing what's running, and trimming the bill without trimming reliability.
Performance optimization. Redis caching, Sidekiq background processing, query optimization — the work that turns a sluggish app into a fast one.
Integrations that don't break. Payments (Stripe, PayPal, PCI compliance, fraud detection), calendars and conferencing, and the retry logic that keeps them stable.
Leading teams. Technical strategy, architecture decisions, code review culture, and working across functions to actually ship.
E-commerce · Healthcare · Education · Real Estate · Scheduling & Productivity SaaS
- Bachelor's in Computer Science
- AWS Certified Solutions Architect – Professional
Firm believer that communication is key — most engineering problems turn out to be conversation problems wearing a trench coat. Off the keyboard you'll find me at a table tennis table or somewhere with good coffee.
Open to interesting problems — especially AI systems that need to survive contact with production, and Rails at scale.