Igor Tamm

I’m a senior/lead backend engineer specializing in Go, building high-throughput payment systems and real-time messaging platforms on PostgreSQL, Kafka, gRPC, and Kubernetes. I focus on correctness, performance, and operational excellence—so services stay fast, reliable, and observable under load.

At Small & Skif, I led backend for a real-time corporate messenger (~15K MAU) from scratch, owning end-to-end architecture using Clean Architecture and ~11 bounded-context microservices. I scaled live messaging to 1–5K concurrent WebSocket connections with sub-250ms fan-out using Centrifugo and Redis pub/sub, sustaining 99.9% uptime and sub-100ms API p99.

I built exactly-once cross-service event flows using a Kafka transactional-outbox, plus PostgreSQL tuning (indexing, connection pooling, idempotent writes). I also delivered fine-grained authorization with Keycloak + Permify RBAC and a full observability stack with OpenTelemetry, Prometheus, and Grafana to reduce incident MTTR.

Previously at T-Bank (formerly Tinkoff), I built and operated high-load Go payment and acquiring services handling thousands of TPS with idempotent gRPC APIs and exactly-once Kafka pipelines, including reconciliation and failure-recovery flows engineered for financial correctness and auditability. I also build self-hosted LLM-inference and RAG infrastructure, including multimodal and cost-aware LLM routing pipelines.