Mikita K.

I architect, scale, and rescue high-load Python backend systems. As an Expert-Vetted Python Architect (Top 1%), I partner with engineering teams to resolve deep performance bottlenecks in applications that have outgrown their initial architecture. With a decade of engineering experience, my work focuses on structural solutions rather than patching symptoms. I build scalable backends capable of handling severe traffic spikes, supported by rigorous database tuning and event-driven design. ENGINEERING FOCUS - High-Load Backend Architecture I design async-first FastAPI and Django services built for scale. I implement horizontal scaling, robust caching layers (Redis), and reliable message brokers (Kafka, RabbitMQ, Celery) to ensure system stability under unpredictable loads. - PostgreSQL Performance & DBA I specialize in advanced database optimization. When standard indexing is no longer sufficient, I implement table partitioning, logical replication, PgBouncer connection pooling, and complex query refactoring. My priority is redesigning data access patterns and eliminating N+1 anomalies at the ORM layer. - System Refactoring & Technical Debt I regularly audit struggling backends. I provide comprehensive profiling for memory leaks, slow queries, and I/O bottlenecks. I execute phased refactoring without halting production, delivering fully tested, documented code integrated into your CI/CD pipelines. • Production-Grade AI Integration Integrating OpenAI/Anthropic APIs into backend workflows. I build fault-tolerant RAG pipelines using pgvector, ensuring strict rate-limiting, cost-monitoring per token, and fallback mechanisms for predictable enterprise use. RECENT INFRASTRUCTURE WINS - Resolved critical database lockups in a monolithic Django app by implementing PgBouncer connection pooling and refactoring ORM queries, reducing required compute infrastructure by over 40%; - Redesigned data access patterns on a 2.5B+ row PostgreSQL table via declarative partitioning and materialized views, dropping complex analytical query latency from ~35s to under 250ms; - Migrated a high-traffic synchronous Django API to async FastAPI, reducing average response times by 70% with zero downtime. Due to the nature of enterprise work, client names are strictly confidential. I am happy to discuss specific architectural patterns and system designs during an interview. Tech Stack: Python, FastAPI, Django, PostgreSQL, ClickHouse, Redis, RabbitMQ, Kafka, Celery, Docker, Kubernetes, AWS, GCP. I am currently accepting select engineering projects. Let's schedule a technical sync to discuss your current architectural bottlenecks.