Alikhan Sainov

Senior Machine Learning Engineer / Acting ML Practice Lead

GDPartners

Leads applied AI work and builds production automation and legal AI systems, with reported improvements in matching accuracy and recall.

• Leading applied AI and GenAI initiatives across enterprise automation, computer vision, and legal AI domains. • Redesigned a production invoice automation platform (React, FastAPI, PostgreSQL) in under 2 weeks supporting regional operations across multiple countries (IMT / Wolt). • Improved invoice-to-PO semantic matching accuracy from 50% to 85% using embedding-based retrieval and semantic ML pipelines. • Increased matching recall from 28% to 93% by redesigning temporal filtering and candidate retrieval logic.

AI Engineer

Perseptex

Built healthcare and enterprise AI products, including voice workflows, RAG assistants, and scalable LLM inference APIs.

• Designed and deployed ‘Xerxes”, an LLM-powered healthcare call-quality platform for patient issue detection and escalation workflows. • Co-developed ‘Zure”, integrating pVerify APIs with Gemini-based agents for automated insurance verification and structured claim analysis. • Built multilingual voice AI pipelines using Gemini Live and Telnyx with voicemail detection, intent extraction, and real-time patient interaction workflows. • Developed RAG-based enterprise assistants using Gemini, Milvus, and PostgreSQL for internal knowledge retrieval and HR automation.

Machine Learning Engineer

Kaspi.kz

Kazakhstan digital platform offering payments, marketplace, and financial services.

Improved LLM-based product matching across millions of catalog items, raising recall from 20% to 78% and accuracy from 70% to 91%.

• Fine-tuned LLM-based product matching systems across the Kaspi Marketplace catalog (millions of SKUs), increasing recall from 20% to 78% and accuracy from 70% to 91%. • Built LangChain-based pipelines for automated product attribute extraction and catalog enrichment using LLMs. • Developed multiprocessing workflows for large-scale product matching using Polars for memory-efficient data processing. • Built data pipelines for training data preparation, evaluation, and production ML workflows. • Worked with production ML systems serving millions of daily active users.

Senior Associate

PwC Luxembourg

Luxembourg member firm of the global professional services network PwC.

Led financial audits and analyzed complex client data, including anomaly detection, reporting, and risk assessment.

• Led audits of financial statements under IFRS and Lux GAAP with a focus on data-driven analysis and risk assessment. • Performed large-scale financial data analysis and anomaly detection across complex client datasets. • Worked with financial instruments, consolidations, and multi-entity reporting structures. • Built a strong foundation in financial systems, compliance processes, and enterprise document workflows.

Machine Learning Engineer (Intern)

Astana Hub

Technology park and startup ecosystem in Kazakhstan.

Built a crime prediction model reported at 96% accuracy and developed sentiment analysis tools.

• Developed a crime prediction model with 96% accuracy in collaboration with the Prosecutor General’s Office. • Built sentiment analysis tools using deep learning and Python libraries.

Graduate Coursework, Applied Information and Data Science

Lucerne University of Applied Sciences and Arts (HSLU, Switzerland)

Master of Science, Finance

Budapest University of Technology and Economics

Bachelor of Business Administration

National Louis University