Yernar Bissultanov

AI Engineer

ABR tech

Built personalized search and recommendations, a customer analytics platform, and internal AI tools.

• Designed and built the search and recommendation system for a food-delivery and lifestyle mobile app, personalizing search and content discovery for each user and surfacing relevant cross-sell recommendations. • Designed and built the company’s customer-analytics platform, enabling data-driven, segment-level marketing instead of intuition-based decisions. • Developed in-house AI automation and agentic tools, streamlining analytics and operational workflows across the team.

Gen AI Engineer

Dualboot Partners

Software development and technology consulting company.

Led production RAG development, reduced escalated tickets by 43%, and cut LLM API costs by $5,000 per month.

• Led development of a production RAG system for client chatbots and internal assistants, reducing escalated tickets by 43% and cutting average response time. • Built a multimodal semantic chunking framework for context-aware document segmentation and improved ranking in RAG pipelines. • Fine-tuned LLaMA with PEFT for product matching, reducing API costs by $5,000 per month.

Gen AI Engineer

Kaspi Bank

Kazakhstan bank and financial technology company.

Improved recommendation and LLM search systems, and reduced annotation time by over 40%.

• Integrated NLP modules into the recommendation engine, achieving 0.83 NDCG and improving product discovery for millions of customers. • Built multilingual data-chunking pipelines, boosting search accuracy of LLM-based tools. • Automated pre-labeling workflows, cutting annotation time by 40%+ and accelerating model release cycles.

Senior Data Scientist / Team Lead

Halyk Bank

Kazakhstan commercial bank.

Led three data scientists and delivered credit and campaign models, including a credit approval model that increased conversion by over 3x.

• Developed and deployed a CatBoost-based credit approval model that increased customer conversion by over 3x while maintaining risk profile targets. • Built and executed A/B testing frameworks, informing product strategy and increasing campaign engagement by 60%.

Machine Learning Engineer

Alem Research

Built and deployed NLP and churn models, including a classifier with 91%+ F1 and automated processing of 70,000+ feedback items monthly.

• Implemented a toxic-language detection model that improved platform health and user trust, contributing to increased user retention and subsequent sales growth. • Fine-tuned and deployed BERT models for classification and NER, automating processing of more than 70,000 feedback items monthly. • Delivered a churn classifier with 91%+ F1, enabling proactive retention campaigns.

Bachelor, Computer Science

Kazakh-British Technical University

Master, Computer Science

Kazakh-British Technical University