Swarnnika Raj Singh
Building cloud platforms for the AI era.
Trusted By
IBM Software
CurrentHashiCorp Cloud Platform Services
Platform Engineer II
UST
Software Development Engineer III · Platform
HealthWorksAI
Software Development Engineer II
Sarthi AI
Founding Engineer · Platform
Omnicell
Software Development Engineer I
What I Build
Grouped by the problem, not the stack.
Platform Engineering
Teams lose weeks to inconsistent environments and manual provisioning. I build internal platforms and golden paths that let engineers ship without waiting on infrastructure.
AI Infrastructure
Running AI workloads reliably needs more than an API key. I design vector database platforms, inference pipelines, and control planes that keep AI systems fast, observable, and cost-aware.
Developer Platforms
Deployment should feel invisible. I build CI/CD and self-service deployment systems modeled on Vercel-style workflows, so shipping code is a non-event.
Cloud Automation
Manual cloud operations don't scale past a handful of services. I automate provisioning, cost controls, and infrastructure lifecycle with Terraform and event-driven pipelines.
Distributed Systems
Systems fail in ways single machines never do. I design for partial failure — queues, retries, idempotency, and consistency guarantees that hold up under real load.
High Performance APIs
Latency compounds across a system. I build gRPC and REST services in Golang tuned for throughput, backed by caching and queueing strategies that keep tail latency low.
Case Studies
Engineering thinking, not screenshots.
Engineering Philosophy
How I think about building systems.
Scalable software is a people problem before it's a systems problem
Most systems don't fail at the architecture diagram — they fail because three teams built around different assumptions. I design interfaces and ownership boundaries before I design services.
Kubernetes is not always the answer
If a team can't operate what they run, orchestration complexity becomes a liability, not a capability. I reach for Kubernetes when the problem is genuinely distributed — not by default.
Simplicity is a discipline, not a starting point
Simple systems are the result of cutting things away under pressure, not the absence of effort. Every abstraction I add has to earn its place against the cost of understanding it.
Infrastructure should empower developers, not gatekeep them
A platform's job is to make the right thing the easy thing. If developers route around your platform, the platform — not the developer — has failed.
Reliability is a feature you design for, not one you patch in
High availability, retries, and graceful degradation are architectural decisions made at design time. Bolting them on after an outage is always more expensive.
Writing
Things worth writing down.
No generic how-tos. Writing focused on production lessons from building platforms and distributed systems.
Open Source
Pinned repositories & experiments.
AI-assisted Kubernetes optimization control plane — workload classification, explainable recommendations, and pre-action simulation.
Python client SDK for SwarnDB, a high-performance vector database — connection pooling, a chainable query builder for dense, sparse, and tensor search, and DataFrame result conversion.
Metered memory service for RAG and AI agents — FastAPI, Redis-backed rate limiting, and PostgreSQL-backed balance and transaction tracking.
Event-driven video transcoding pipeline using AWS Lambda, ECS, and Redis for queueing and pub/sub coordination.
Self-service deployment platform for static frontends with queued builds and live log streaming.
Multi-user WebRTC video calling infrastructure with WebSocket-based signaling.
Real-time chat platform with face-recognition auth and speech-to-text, built on AWS Rekognition.
Roadmap
- Infrastructure templates for platform teams standing up internal developer platforms
- A minimal Golang toolkit for control-plane style services
- Writeups extracted from private production systems, generalized into reusable patterns
Beyond Code
A few things that matter just as much.
Animal rights matter to me like family, not a side interest. Beyond that: veganism, fitness, and generative art.
Animal Rights
I co-run Uddharini Foundation with a small team — rescuing animals in need, running ahimsa and veganism classes for schools, and working toward building a stray animal hospital.
Veganism
Veganism isn't a diet — it's a rejection of animal exploitation in food, clothing, and every other form. A daily practice of that belief, not a trend.
Generative Math Art
Algorithmic art built from trigonometric and geometric equations, not photographs — collected as Magic of CosTheta.
Fitness
Strength training as a discipline — consistency over intensity.
Contact
Available For
Platform Consulting
Advising teams on internal platform strategy, developer experience, and infrastructure roadmap.
Architecture Reviews
Second opinions on distributed systems, control plane design, and scalability tradeoffs before they become production incidents.
Technical Mentorship
Working with engineers moving from application development into platform and infrastructure engineering.
Speaking
Talks and panels on platform engineering, distributed systems, and AI infrastructure.
Open Source Collaboration
Contributing to or collaborating on infrastructure tooling, control planes, and developer platform projects.
Startup Advisory
Early-stage technical guidance for founders building platform or infrastructure-heavy products.