Наталья Сажина

Exchanges data processing team. Created and maintained various Go microservices in highly distributed and mission-critical market data systems at Tinkoff Investment (3 projects with over 95 services).

- Performing all phases of software development including analysis, design, implementation, system integration and QA support.

- Сode reviewing and setting up best practices within the team for design and development.

• Reduced the cost of the production resources by a factor of two by breaking down a streaming monolith application into microservices.
• Built a library with gRPC bidirectional stream functionalities used across all our micro-services.
• Developed high-load tests to benchmark the performance of distributed streaming applications. Optimized resource utilization and decreased API latency. Developed autotests to check subscriptions consistency of a pub-sub system.
• Implemented multi-level unit and integration tests following TDD, BDD approaches.
• Migrated several services to newer infrastructure. Developed CI/CD for numbers of projects.
• Actively participated in code reviews. Cultivated development processes (code review, CI/CD, tasks decomposition, TBD workflow, code coverage increasing), engineering culture. Promoted good software engineering practices: TDD, BDD, DDD, SOLID, KISS, DRY, YAGNI, BDUF, APO. Practiced pair programming approach - increased features implementation by 20-30%.
• Significantly reduced the "Bus Factor" via increasing documentation volume by ~40% and establishing an on-call practice.
• Increased engineering collaboration by encouraging frequent discussions and constant feedback from team members.
• Helped the SRE team with learning Golang, and taught best practices.

• Designed and implemented a gRPC bidirectional stream micro-service from scratch with more than 90% test coverage.
• Redesigned multiple backend components to resolve technical debts. Fixed mission-critical bugs involving race conditions, synchronization and multi-threading.
• Significantly increased performance of 30+ services by adding in-memory private/shared caching. Carried out performance testing and usage analysis, provided strong/eventual data consistency (including high dynamic data). Managed concurrency, supported cache failover, and effective utilization on heavy loads.
• Setup monitoring and build alert rules to recognize application crashes, inconsistent states and project failures. Increased alert rules count by x5. Developed various troubleshooting scenarios.
• Increased logs informativeness. Reduced load on log export system by removing bugs that generated 99% rubbish logs of API (200 000l/min → 2000 l/min per pod).
• Migrated services of 3 projects to newer infrastructure: Rancher → Kubernetes.
• Made DBMSes comparison research as a part of the data migration task. Found the core performance problem and resolved it with instruments of current DBMS. Affected queries were boosted by x3+. High-cost migration process was stopped.
• Increased knowledge sharing and total technical documentation volume. Updated half of the technical documentation volume of 3 projects.
• Applied test design practices to write tests faster with more efficiency. Set up mock-testing practice in a team.
• Сarried out the onboarding process of Software Engineers.