CEO & Founder at Humarin (JSC Explat)
Moldova
Vladimir
Vorobiev
Council Member
Vladimir Vorobiev is a CEO of Expert Platforms, a joint-stock company where he leads the development and implementation of advanced AI solutions for automating medical processes during patient consultations, including clinical decision support systems. Based in Russia, he brings four years of machine learning and deep learning experience combined with one year of backend development expertise to healthcare technology innovation.
Currently serving as CEO and Founder of Humarin (JSC Explat) since February 2025, Vladimir Vorobiev designed and implemented the core platform for building clinical agents based on large language models. He created an RPA system for streamlined hospital information system integration, assembled and managed a cross-functional team, and established the company's product vision and development milestones. Under his leadership, Humarin launched pilot studies at major hospitals including Fomina Clinic and Sechenov University Clinic, secured the company's first commercial contract, and registered intellectual property for an AI-powered clinical decision support system. The company focuses on building an AI platform that reduces time doctors spend on routine tasks by up to 70%.
From December 2023 to April 2025, Vladimir Vorobiev worked as an NLP Data Engineer at Sberbank in Moscow, where he built legal-domain large language model pipelines for GigaLegal, automating legal workflows in SberLegal and corporate departments. He developed advanced legal search and retrieval pipelines, created a situation clarification module for initial client consultations, and achieved 70% quality metrics on mid-to-hard legal queries. The solution was successfully sold to the Ministry of Justice of the Amur Region, with pilots conducted for major corporations.
Prior to that role, Vladimir Vorobiev served as a Data Scientist at Sberbank from September 2023 to December 2023, conducting applied research under non-disclosure agreements and focusing on advanced machine learning applications. His first position at Sberbank was as an Intern Data Engineer from July 2022 to November 2022, where at age 17 he built a machine learning library for automated question-answering over internal legal and governmental documents.
In 2024, Vladimir Vorobiev developed a lung disease diagnostic system capable of diagnosing viral pneumonia in eight seconds with 100% accuracy while demonstrating effectiveness in detecting other lung diseases. This clinical decision support system processed electronic medical records to deliver preliminary and final diagnoses, achieving the highest score across six lung disease categories with an average score of 0.87. The system won first place in the AI'm Doctor competition, organized by the National Technology Initiative Foundation in partnership with the Skolkovo Foundation and Sechenov University, earning him 35 million rubles in prize funding for developing the best clinical decision support system for lung diseases in Russia.
Currently serving as CEO and Founder of Humarin (JSC Explat) since February 2025, Vladimir Vorobiev designed and implemented the core platform for building clinical agents based on large language models. He created an RPA system for streamlined hospital information system integration, assembled and managed a cross-functional team, and established the company's product vision and development milestones. Under his leadership, Humarin launched pilot studies at major hospitals including Fomina Clinic and Sechenov University Clinic, secured the company's first commercial contract, and registered intellectual property for an AI-powered clinical decision support system. The company focuses on building an AI platform that reduces time doctors spend on routine tasks by up to 70%.
From December 2023 to April 2025, Vladimir Vorobiev worked as an NLP Data Engineer at Sberbank in Moscow, where he built legal-domain large language model pipelines for GigaLegal, automating legal workflows in SberLegal and corporate departments. He developed advanced legal search and retrieval pipelines, created a situation clarification module for initial client consultations, and achieved 70% quality metrics on mid-to-hard legal queries. The solution was successfully sold to the Ministry of Justice of the Amur Region, with pilots conducted for major corporations.
Prior to that role, Vladimir Vorobiev served as a Data Scientist at Sberbank from September 2023 to December 2023, conducting applied research under non-disclosure agreements and focusing on advanced machine learning applications. His first position at Sberbank was as an Intern Data Engineer from July 2022 to November 2022, where at age 17 he built a machine learning library for automated question-answering over internal legal and governmental documents.
In 2024, Vladimir Vorobiev developed a lung disease diagnostic system capable of diagnosing viral pneumonia in eight seconds with 100% accuracy while demonstrating effectiveness in detecting other lung diseases. This clinical decision support system processed electronic medical records to deliver preliminary and final diagnoses, achieving the highest score across six lung disease categories with an average score of 0.87. The system won first place in the AI'm Doctor competition, organized by the National Technology Initiative Foundation in partnership with the Skolkovo Foundation and Sechenov University, earning him 35 million rubles in prize funding for developing the best clinical decision support system for lung diseases in Russia.
With the prize funds, Vladimir Vorobiev founded the startup Humarin, where he led development of two modules: Diagnose&Treat (currently under development) and Transcribe. Transcribe functions as an audio prompter that automates the collection of patient complaints, medical history, and physical examination data, records doctor-patient dialogue, and structures the information within medical information systems. Both modules underwent commercial validation and were presented at major healthcare business forums including CIPR, the St. Petersburg International Economic Forum, and the Healthcare Industry Competence Center. Pilot trials of the Transcribe module were conducted at Goryachy Klyuch City Hospital, Veramed Clinic, and Fomina Clinic, with testing at Fomina Clinic confirming that automation of complaint collection and dialogue documentation enabled 50% faster completion of appointment documentation compared to manual entry without sacrificing quality of medical records.
In 2023, Vladimir Vorobiev contributed to development of a text paraphrasing model based on T5 architecture. He created a dataset of 12.6 million training pairs using GPT-based synthetic data and published the model on Hugging Face, where it has been downloaded over 2.9 million times and integrated into Spark NLP and IBM RADAR products.
Vladimir Vorobiev achieved international recognition in artificial intelligence through multiple competitions and programs. He earned first place among five finalists in the AIIJC 2021 NLP track, placed fifth out of 49 participants in the National Technology Olympiad 2023 AI track, and won first place out of 30 teams in the Applied AI Challenge 2023. He also secured second place out of 60 participants in the AI Academy Hackathon 2022. Additionally, he served as a speaker at Yandex Data DoJo and contributed to industry media, with his latest article published in CNews devoted to the integration of artificial intelligence into medicine.
In 2023, Vladimir Vorobiev contributed to development of a text paraphrasing model based on T5 architecture. He created a dataset of 12.6 million training pairs using GPT-based synthetic data and published the model on Hugging Face, where it has been downloaded over 2.9 million times and integrated into Spark NLP and IBM RADAR products.
Vladimir Vorobiev achieved international recognition in artificial intelligence through multiple competitions and programs. He earned first place among five finalists in the AIIJC 2021 NLP track, placed fifth out of 49 participants in the National Technology Olympiad 2023 AI track, and won first place out of 30 teams in the Applied AI Challenge 2023. He also secured second place out of 60 participants in the AI Academy Hackathon 2022. Additionally, he served as a speaker at Yandex Data DoJo and contributed to industry media, with his latest article published in CNews devoted to the integration of artificial intelligence into medicine.