Nazira Kaibalina

Backend Engineer

Jobescape

Built backend infrastructure and AI features for an educational platform, including a personalized AI mentor and course content generation.

• Built backend infrastructure for an AI-powered educational platform from scratch. • Developed an agentic AI mentor with multi-step reasoning and tool use for personalized learner guidance. • Built an AI content generation service for automatic course material creation.

AI Engineer

Nomad Tech

Built an LLM pipeline for a government judicial system that classified offenses and generated indictment documents, cutting preparation time by 27%.

• Designed and implemented an end-to-end AI service for a government judicial system. • Built an agentic LLM pipeline that classifies criminal offenses and auto-generates court-ready indictment documents from unstructured legal evidence, reducing document preparation time by 27%.

Software Engineer

Mind AI @ OurHome & Hanhwa

Built CP-SAT meal planning software and a recommendation model for catering facilities, improving customer satisfaction by 15%.

• Built meal planning optimization software for catering facilities using constraint programming algorithms (CP-SAT). • Implemented and trained recommendation model to customize the generated meal plans and improved customer satisfaction by 15%.

Software Engineer

Your Lab

Built an LLM car sales chatbot and license verification pipeline, reducing manual verification time by 23%.

• Developed an LLM-based car sales chatbot and built an CI/CD pipeline for it. • Implemented a custom RAG pipeline using a vector database to process and respond accurately across 15 vehicle manuals. • Built an automated driver license verification pipeline using LLMs, reducing the manual verification time by 23%.

Machine Learning Engineer

Mind AI

Developed multilingual NLP and chatbot systems, with reported results including 99% pipeline precision and improved customer interaction quality.

• Trained a multilingual NLP pipeline (tokenizer, lemmatizer, tagger, parser) with 99% precision, covering low-resource languages. • Built Korean word segmentation and typo-handler models, improving customer interaction quality by 12%. • Fine-tuned LLMs for intent classification, sentiment analysis, and entity recognition. • Developed a web application for chatbot fallback management, improving chatbot accuracy by 9%.

Master of Engineering, Computer Science and Engineering

Kyung Hee University

Bachelor of Science, Computer Science

Nazarbayev University

Foundation Program

UCL