AI Engineer building production-ready systems that solve real problems.
Strong backend foundation with FastAPI, Django, and scalable APIs.
Focused on agentic AI — RAG pipelines, vector search, and LLM guardrails at JSC Garantbank.
Curious by default — I dig deep until I truly understand how things work.
skills
LangchainExperiencedDockerSkillful
FastAPIExperiencedGraphQLSkillful
SQLExperiencedQdrantSkillful
experience
Agentic AI Engineer now
JSC Garantbank · Tashkent
Dec 2025 — Present
Built an enterprise RAG assistant for bank employees by integrating vector search + cross-encoder reranking + LDAP auth, enabling instant AI-powered Q&A over internal documents.
Launched Garantbank's first AI chatbot on Telegram by building a multi-language RAG pipeline with LangChain + LLM, automating customer support inquiries.
Engineered a multi-format document ingestion pipeline with OCR fallback, powering semantic search across all knowledge sources.
Integrated LLM tool-calling for automated CRM lead capture, detecting purchase intent and validating contacts directly within chatbot conversations.
AI/ML & Software Engineer
Tashkent State Technical University · Tashkent
May 2024 — Dec 2025
Processed patient data, managed model inferences, and returned diagnostic results, ensuring seamless communication with the frontend via AJAX and real-time updates through WebSocket.
Integrated Celery for asynchronous background processing of model inferences, enabling the system to handle 5,000+ requests/month without performance degradation.
Achieved 93% accuracy for the DICOM-based detection model and 96% accuracy for the lung sound detection model, enabling early-stage detection of lung diseases.
Software Engineer
JSC Asakabank · Tashkent
Jun 2023 — Nov 2025
Built a universal calculator for products, including integrated insurance cost computation, used in 100% of loan applications across 12 branches.
Implemented a Lead-to-Client tracking system that improved sales funnel visibility and increased conversion rates by 18% in the first 3 months.
Developed a creditworthiness assessment engine that evaluates loan eligibility based on salary and debt burden, processing over 5,000 applications per month with 96% accuracy.
Designed and integrated a dynamic, role-based permission system supporting 10+ roles and 300+ active users, ensuring secure and flexible access.
Optimized ORM queries with aggregation, annotation, and prefetching, reducing dashboard load time by 85% even on datasets exceeding 1M+ records.
Conducted code reviews and mentored 2 junior developers to improve code quality and ensure adherence to best practices.
Backend Developer
UIC Group · Tashkent
Jun 2023 — Aug 2023
Developed chess platform like chess.com.
Led a team of 4 people and served as a product owner.
Regularly gave talks about clean code, automated tests and best practices.
Python Instructor & Mentor
PDP Academy · Tashkent
Aug 2021 — Jun 2023
Taught and explained advanced backend topics such as Django + AWS, focusing on deployment strategies, environment configuration and cloud-based scalability.
Developed and delivered hands-on sessions on Django + GraphQL, guiding students through schema design, resolver implementation and efficient querying techniques.
Covered real-time communication using Django Channels + WebSocket, demonstrating how to build chat apps and live notification systems.
Explained multi-language support in web applications through Django Localization and Internationalization including language switching, translation files.
Provided deep insights into Django REST Framework (DRF), covering serialization, authentication and building scalable RESTful APIs with best practices.
education
Master, Tashkent University Information Technology
Tashkent
Sept 2021 — Jul 2023
Researched Machine Learning, developing efficient models for pattern recognition, prediction and deep learning.
Delivered talks and mentorship on software design, scalable architecture, and clean code for bachelor's graduates.
Courses: Cloud Platforms, Big Data, Deep Learning
Bachelor, Tashkent University Information Technology Fergana Branch
Fergana
Sept 2017 — Jul 2021
Participated in a C++ competition among 300 students from all branches and took 5th place.
2nd place in the "Most Complex Object" nomination using OpenGL (14 teams participated)
Designed and developed the backend architecture for the platform, ensuring high scalability, security, and maintainability.
Developed and integrated a secure and efficient payment system, allowing seamless financial transactions within the platform.
Implemented a proposal-driven workflow where freelancers submit detailed proposals to employers, outlining their expertise, pricing, and project approach. Designed a ranking and filtering system that allows employers to evaluate proposals based on quality, relevance, and freelancer ratings.
Built a flexible and optimized API using GraphQL, enabling efficient data retrieval and minimizing over-fetching or under-fetching of information. Designed a structured schema and resolvers to enhance query performance, reduce response times, and improve developer experience.
Activity Tracker
JSC Asakabank
Jul 2023 — Sept 2023
The Activity Tracker is a comprehensive system designed to monitor and record computer usage and activity patterns.
The system is built using Python and Django, with a frontend component developed using Tkinter for the GUI. It also utilizes web API for data transmission.
The project demonstrates proficiency in Python programming, GUI development, data serialization and HTTP communication.
It is a valuable tool for individuals interested in self-monitoring computer usage or analyzing productivity patterns.