Denys Shtoma · Tech Lead · Backend Consultant
I lead engineering teams and design backend architectures in Python and Go — payment platforms, high-load APIs, and the databases underneath them. Available for consulting engagements and open to interesting projects.
# ADR-0042: Payment cascade routing
Context: single-provider routing fails during provider outages; conversion drops on declined transactions.
Decision: weighted cascade across providers — automatic retry, idempotent callbacks, per-currency fee rules.
Consequences: higher acceptance rate, no manual retries, fully auditable flow.
About
Based in Kyiv, working remotely with teams worldwide. I've grown from hands-on backend engineering into technical leadership — and I'm building toward the architect role deliberately: ADRs, system design reviews, and owning decisions end to end, not just writing code that works.
My approach: understand the business problem first, then engineer the simplest architecture that solves it. No over-engineering, no hype-driven development. I stay close to the details that decide whether a system survives — query plans, failure modes, migration strategies — while keeping the team shipping.
Hexagonal architecture, DDD, event-driven systems. Decisions written down as ADRs, not tribal knowledge.
Mentoring engineers through code review, pair programming, and design sessions. Coverage gates and review culture that stick.
PostgreSQL performance is a specialty: execution plans, indexing, partitioning, zero-downtime migrations on hot tables.
PCI-relevant flows, HMAC-signed webhooks, card-data masking, 3DS. Payments taught me to treat security as a requirement, not a feature.
Selected Work
A cross-section of production work — described by the problems solved, not the job titles held.
Own the provider-integration layer of a payin/payout platform connected to 170+ payment providers: webhook processing with HMAC verification, 3DS flows, cascade routing across providers, and multi-currency fee calculation with correct minor-unit handling.
Diagnosed and removed PostgreSQL bottlenecks on hot transaction tables: execution-plan analysis, composite and partial indexes, range partitioning, and zero-downtime online migrations — cutting worst-case latency from seconds to tens of milliseconds.
Built Go and Python microservices for real-time ingestion and processing — concurrent pipelines handling 10K+ requests per second with backpressure, graceful shutdown, and versioned APIs that let clients migrate without breaking.
Developed and scaled REST APIs serving 5M+ daily requests for enterprise clients: Kafka-based data pipelines, Elasticsearch full-text search across microservices, and Redis caching strategies that cut response times by 40%.
Delivered event-driven, serverless backends on AWS — Lambda services, Kafka and RabbitMQ pipelines, containerized deployments — designed for scalability with infrastructure cost kept under control.
Consolidated scattered legacy workflows into a unified internal platform: Django services with server-rendered UI, asynchronous Celery pipelines with idempotent task handling — cutting time-to-ship for internal tools by ~40%.
Services
Consulting engagements first — audits, architecture, leadership. And when a project is genuinely interesting, I build.
Architecture audits, system design, and technology selection. A second opinion before a costly decision — or a written ADR trail after it.
High-load APIs, microservices, and payment integrations in Python and Go. REST, gRPC, and event-driven systems built for scale and reliability.
PostgreSQL tuning as a dedicated engagement: slow-query diagnosis, indexing and partitioning strategy, and zero-downtime migrations on live systems.
Fractional tech lead for teams that need senior direction: review culture, delivery process, mentoring, and engineering standards that outlast me.
Understanding your goals, constraints, and existing infrastructure.
Designing the solution, choosing the stack, writing the decisions down.
Building iteratively with regular demos, reviews, and feedback loops.
Deployment, documentation, knowledge transfer, and ongoing support.
Tech Stack
Contact
An architecture review, a performance problem, a backend built from scratch, or a second opinion before a big decision — describe it in a few sentences and I'll reply within one business day.