Venture Fellow (Undergraduate Research Scholar), NanoTRIZ Innovation Institute
Mr. Daulet Imangaliyev
Venture Fellow (Undergraduate Research Scholar), NanoTRIZ Innovation Institute
Current Affiliation: University of Bologna, Italy
Research Focus: Developing LLM for Scientific Research / Automating the research process
Available for supervision
Projects for students: Ad-hoc, fine-tuning LLMs and speech models to automate scientific research workflows
Biography:
Daulet Imangaliyev is a high-capability technical researcher and Venture Fellow (Builder Track) at the NanoTRIZ Innovation Institute. His career is characterized by a rapid trajectory through international academic environments and a focused expertise in the development of Large Language Models (LLMs) and speech-based Artificial Intelligence. Daulet has participated in prestigious research and exchange programs across the globe, including appointments and coursework at the University of Bologna, Hochschule Schmalkalden, the Beijing Institute of Technology, and the University of Birmingham. His professional journey reflects a dedicated commitment to bridging the gap between statistical mathematics and scalable AI engineering.
Following his selection into the highly selective Nfactorial Incubator, where he was chosen from a pool of over 4,000 applicants. Daulet demonstrated a unique ability to translate complex algorithms into functional applications. This excellence led to his appointment at the Institute of Smart Systems and Artificial Intelligence (ISSAI), where he served as an Undergraduate Research Assistant. At ISSAI, he was instrumental in fine-tuning state-of-the-art multilingual models, such as Meta’s MMS and OpenAI’s Whisper, significantly improving performance metrics for 22 languages. This international foundation now serves as the basis for his work at NanoTRIZ, where he contributes to the co-development of proprietary AI architectures and automated research workflows.
Research Interests:
Daulet Imangaliyev is establishing himself as a specialist in autonomous research systems and multilingual speech processing. His work at NanoTRIZ focuses on the development of an LLM-driven platform designed to automate the entire scientific research process, from literature synthesis to automated prototyping.
Core research topics and technical interests include:
Scientific LLM Automation: Developing internal Institute tools for mapping research gaps and capability analysis.
Multilingual Speech AI: Fine-tuning large-scale speech models (Whisper/MMS) and optimizing NLU for diverse user scenarios.
Scalable AI Engineering: Optimizing high-performance computing environments and reducing RAM usage in model training through multi-stage Docker builds.+3
Parallel Web Crawling: Engineering high-speed data collection systems, having successfully collected over 1TB of textual data from 200,000+ URLs.
Statistical Modeling: Applying advanced inference, econometrics, and machine learning to financial and scientific datasets.
As a Venture Fellow, Daulet also acts as a Technical Mentor, providing ad-hoc guidance and code reviews for Junior Research Fellows