Alisher Zhanayev

Designed and maintained backend services for a production platform, focusing on reliability, scalability, and data consistency. Implemented testing strategy raising test coverage from 0% to 80% in 6 months, reducing production regressions. Built Elasticsearch-based search to improve data retrieval performance. Implemented Grafana-based monitoring and alerting for proactive incident detection. Improved data validation and transaction monitoring, reducing fraud-related incidents by 5%. Designed and implemented a production-ready LLM-powered assistant for internal workflows, including an end-to-end LLM pipeline with preprocessing, RAG using a vector database, and post-processing. Added a secondary LLM verification layer to assess correctness and relevance before returning results. Implemented LLM observability and evaluation by logging full request lifecycles for debugging and iterative improvements.

Senior Python Developer focusing on backend reliability, scalability, and data consistency. Designed and implemented a production-ready LLM-powered assistant, built an end-to-end LLM pipeline with Retrieval-Augmented Generation (RAG), introduced a secondary verification layer to ensure output correctness, and established full lifecycle observability for debugging and quality analysis. Implemented Elasticsearch-based search, Grafana-based monitoring/alerting, and data validation/transaction monitoring to reduce regressions and fraud.

Improved ETA prediction accuracy from 84% to 91% by optimizing routing algorithms and backend logic. Decomposed a large monolithic system into 8 Go-based microservices, improved scalability and reduced resource usage. Increased automated test coverage from 90% to 95%, reduced production failures and regressions. Optimized payment workflows, reducing failed transactions from 6% to 4% and integrated multiple payment service providers to improve reliability.