Senior full stack engineer, five years building the unglamorous layer underneath products that can't fall over: identity, ingestion, delivery pipelines and the dashboards on top of them.
Tenant isolation, roles, permissions and audit trail were an afterthought — and the auth layer is now your bottleneck.
Consumer lag, duplicate processing, silently dropped messages, and no dead-letter story when things go wrong.
Notifications, jobs and integrations break and nobody finds out until a customer does.
No repeatable deploy, environment config or secret management. Releases need a person and a prayer.
Read the code, map the data and event flow, benchmark the hot paths. We agree on three measurable targets before a line is written.
Hybrid RBAC and tenant isolation, partition-aware consumers, retries / DLQ / idempotency, Redis caching, API hardening.
Helm on AKS through CI/CD, Key Vault secrets, observability with Prometheus, Grafana and OpenTelemetry. Runbooks and a working session with your team.
I work in your repo, your board and your review process — and leave documentation, not dependency.
Auth was hammering PostgreSQL on every request, across every tenant.
Built a multi-tenant IAM service with hybrid RBAC, tenant isolation, audit logging and Redis permission caching, plus 25+ Spring Boot APIs for JWT validation, permission evaluation and session management.
Telemetry ingestion was lagging and double-processing events.
Rebuilt on Azure Event Hub with partition-aware async consumers, batch and checkpoint tuning, DLQ and idempotent consumers — now also delivering 200K+ notifications daily and 300K+ messages a month with zero duplicate ingestion.
Three years running end-to-end builds for logistics operators.
Shipped a lead-to-conversion ERP covering sales, HR, payroll, commission and slab modules with fine-grained RBAC and Razorpay / Axis / ICICI gateways secured by hybrid encryption — plus a VSL landing page on Cloudflare Workers using chunked progressive video.
A platform, pipeline or IAM build taken all the way to production — scope, milestones and price agreed up front.
Architecture, delivery and code review alongside your team, with real commits — not just advice on a call.
An audit of your scaling, auth or event flow, ending in a costed remediation plan you can execute with or without me.
Contract-to-hire considered for teams building long-term platform capability.
A 30-minute call is usually enough to tell you whether this is a two-week fix or a ten-week build — and I'll say so either way.
hello@samyakkhatua.in→