Dr. Dr. Roman Gorbunov

• Evaluation of Large Language Models (LLMs) • Retrieval Augmented Generation (RAG) • Prompt Engineering • Fine-tuning of LLMs • LangChain • Modular Reasoning, Knowledge, and Language Systems for Autonomous Agents (MRKL) • Autonomous LLMs Agents (LLMs + Memory + Tools)

• Leading strategic development of AI platform for an automated generation of trading strategies • Presentation of AI platform to clients, investors and internal users • LLM, GPT, prompt engineering, few-shot learning • Facilitation of communication between Machine Learning Experts, Financial Traders and Software Developers • Design, implementation and evaluation of new Machine Learning algorithms • Development of statistical tests for verification of patterns under a low signal to noise ratio

• Conducting forefront research and development in all aspects of human language technology • Using many real-world data sources to build highly scalable ML solutions • Scaling up machine learning systems to new languages and devices • Supporting new features and locales for which we have limited data • Partnering with science, engineering, data and product teams all over the world • Planning of team work and reporting on team performance • Being a team‘s point of contact for external requests

• Leading a team of data scientists working on development of data driven products • Playing a role of data science product owner within agile methodology (Scrum, KANBAN) • Development of price elasticity and sales cannibalization models for retail industry • Presentation of commercial aspects of data science results to management and internal teams • Discussion of business opportunities with potential clients • Turning the machine learning models into scalable products

• Combining diverse data sets to model dependency of demand on socio-demographics • Analysis of retail data and consumer data • Building of predictive models • Exploratory analysis • Choosing tools for data science infrastructure

• Data Preparation, Data Analysis, Data Visualization, Data Modeling • Machine Learning • Multivariate Time Series Analysis, Time Series Modeling, Time Series Predictions • Predictive Analytics • Supply Chain Management, Inventory Management • Spare Parts Management • Dynamic Pricing • Reinforcement Learning, Optimal Control Theory, Dynamic Programming • Hidden Markov Processes

• Analysis of Behavioral Data • Behavioral Game Theory • Social Preferences and Social Factors • Agent Based Modeling • Data Mining on Behavioral Data • Finding Hidden Factors in Behavioral Time Series • Mathematical Modeling of Cooperation and Conflicts Researcher at ESA MARS-500 Project: http://www.esa.int/Our_Activities/Human_Spaceflight/Mars500. Development of a computer system of an automatic monitoring of emotional states and interpersonal relationships.

• Analysis of Big Biophysical Data • Ab initio descriptions of vibrations in DNA • Calculations of two-dimensional infrared spectra of DNA • Molecular dynamic simulations of large biomolecules • Study of influence of external environment on functionality of biomolecules

• Ab initio parameterization of vibrational Hamiltonians of peptides. • Combination of classical and quantum descriptions of vibrations in peptides. • Study of approximations used in calculations of one-dimensional infrared spectra. • Instructor in practical classes in inorganic chemistry for students. • Instructor in practical classes in quantum mechanics for students.

• Quantum-mechanical calculations of crystals' mechanical properties. • Calculations of electrons binding energies. • Instructor in practical classes in physics for students.