- Implemented new procurement procedures (deposit auctions, cargo service tenders) using an existing state machine framework, focusing on business logic rather than infrastructure. - Wrote validation logic for procedure fields, implemented user notifications, and handled mandatory notifications to authorized government bodies. - Maintained and extended existing procedures (tenders, selections, auctions, ads, requests), including adding new workflow stages and features. - Refactored procedure rendering logic by replacing a large, state-driven “god” module with explicit, procedure-specific view modules, making the rendering logic easier to understand and debug. - Moved field configuration (type, read-only state, enums, layout metadata) out of procedure state and into code, storing only field values in state. This reduced stored procedure state size by ~70% and made workflows easier to reason about. - Fixed bugs primarily related to complex business rules and integrations with internal and third-party services.

- Worked on backend services for a banking application prior to production launch. - Implemented parts of the user authentication flow and APIs used by internal services.

- Built a salary calculation engine for an internal ERP system handling taxes, insurance contributions, and organization-specific rules. - Designed and implemented a formula parser and evaluator in Elixir, supporting user-defined expressions with math and logical operators, brackets, and variable substitution. - Variables were resolved dynamically based on organization, branch, employee, and time context (day/month/year). - Integrated with multiple Elixir and Node.js services via Kafka to fetch required data for calculations. - Services were deployed independently in a cluster and communicated using both Kafka and direct Elixir message passing. - Focused on keeping the engine accurate, flexible, and fast despite complex and evolving rule sets.

- Implemented ATM clustering for an in-app map using PostGIS to group nearby ATMs efficiently. - Built a log batching pipeline for analytics events from multiple services and sent them to ClickHouse using Broadway. - Worked on payday credit features, mainly integrating with the internal clearing department and exposing APIs. - Implemented promotional features highlighting merchants with higher cashback or discounts inside the mobile app. - Contributed to backend services supporting one of the early expense analytics features on the local market.

- Built and maintain a live Elixir service that analyzes in-game item values for 500-1000 users every league. - Polls official game API for stash changes, processes data concurrently using Broadway, and calculates realistic prices while filtering out price-fixing listings. - Designed data pipeline to handle high-volume, real-time updates efficiently with batch writes to PostgreSQL. - Deployed on a VPS using Docker and Dokploy, ensuring uptime and performance over 2 years in production. - Collaborated with a frontend developer for UI/UX; owned all backend, data, and DevOps responsibilities.