name:       Shivansh Shekhar
role:       Backend Engineer (SDE 2)
location:   India · Remote-first
experience: 3+ years

core_stack:
  - TypeScript / Node.js / Fastify
  - PostgreSQL / Drizzle ORM / pgvector
  - Redis · BullMQ · RabbitMQ
  - Docker · GitHub Actions · DigitalOcean / AWS

specialties:
  - Multi-tenant SaaS architectures
  - Distributed systems & event-driven design
  - LLM infrastructure & AI-native backends
  - TypeScript SDKs & developer tooling

currently:  Building in public · targeting async-first product teams internationally
beyond_code: Books · Games · internet lore · Volleyball

One OpenAI-compatible endpoint to rule them all.

A production-grade LLM API gateway unifying OpenAI, Anthropic, Google, Cohere, and Mistral behind a single interface — with semantic caching, RBAC, cost attribution, PII masking, and per-provider circuit breakers.

Fastify TypeScript PostgreSQL + pgvector Redis Docker GitHub Actions

• ~40% reduction in redundant LLM API costs via pgvector HNSW semantic cache
• Tenant-isolated RBAC + hybrid API key management
• Ships with a TypeScript SDK, Next.js analytics dashboard, and full Compose stack

Zero-latency flag evaluation with real-time propagation.

Enterprise-grade feature flagging with a custom recursive rule engine, consistent hashing for sticky rollouts, and SSE + Redis Pub/Sub for instant propagation — no round-trips to the server on every request.

Fastify Turborepo PostgreSQL Redis Pub/Sub SSE TypeScript SDK

• Three-tier SDK fallback logic for zero-latency local evaluation
• Two-level RBAC management dashboard built in Next.js
• Fumadocs documentation site + Docker + automated release pipelines

One schema. Zero drift across DB, API, and docs.

An open-source TypeScript code generation library that produces Drizzle ORM schemas, typed repositories, Fastify validators, and OpenAPI 3.1 specs from a single resource definition.

TypeScript Drizzle Fastify OpenAPI 3.1 CLI

• Multi-adapter architecture: Drizzle · Fastify · OpenAPI generators
• CLI with project scaffolding and granular generation commands
• Eliminates schema drift across database, API, and documentation layers

Languages	Backend & APIs	Databases & Cache
Infra & Cloud	Queues & Distributed	Observability & Tooling

I'm actively looking for remote backend / full-stack roles at product teams and startups — ideally async-first, with real engineering problems to solve. If that sounds like your team, reach out.

I'm also always up to talk distributed systems, LLM infra, TypeScript DSL design, or why The Backrooms is genuinely unsettling.