Description
We need to develop an agent that will orchestrate several ML models to analyze bond order books. Based on this analysis, it will suggest optimal options for buying/selling securities to traders.
Responsibilities
- Gathering requirements for models and communicating with the client and DE
- Defining target metrics for the solution and aligning them with the client
- Data collection and preparation (Hadoop, pyspark, writing parsers for gathering external data)
- Managing Middle DS resources
- Building models for tabular data
- Anomaly detection in time series
- Building time series forecasting models
- Classifying news by its degree of impact on the bond market
- Generating trading signals based on unstructured data
- Creating a RAG database for generating textual justifications of recommendations for traders
- Adapting LLMs to formulate trading ideas in natural language
- Testing various hypotheses.
Requirements
- Data Scientist experience from 2 years
- Degree in "Mathematics", "Physics", "Mathematical Methods in Economics" (preferably graduates of Moscow State University, Moscow Institute of Physics and Technology, Higher School of Economics and other leading universities in the country)
- Deep knowledge of probability theory and mathematical statistics
- Skills in writing high-quality Python code (knowledge of OOP, ability to write code optimized for speed and memory)
- Understanding of how ML model algorithms work (boosting, neural networks, NLP models, LLMs)
- Proficiency in several Python libraries for classical ML from the list or relevant ones (numpy, sklearn, pandas, scikit-learn, matplotlib/seaborn/plotly, catboost/lightgbm/xgboost, fbprophet, pygam)
- Proficiency in pytorch or other DL frameworks
- Hands-on experience with neural network models (RNN, LSTM, Transformer, BERT)
- Experience in developing NLP models (NLTK, spaCy, gensim, transformers)
- Experience using GigaChat, Kandinsky and similar tools in products, skills in creating and using AI agents
- Instrumental mastery of AI for analysis, generation, and automation.
Conditions
- Comfortable modern office near Kutuzovskaya metro station
- Hybrid work format
- Annual salary review. Annual bonus
- Corporate gym and relaxation areas
- More than 400 educational programs from SberUniversity for professional and career development
- Adaptation program and supervisor assistance at the start
- Extended voluntary health insurance (VHI), preferential insurance for family, and corporate pension program
- Flexible mortgage discount equal to 1/3 of the Central Bank's key rate
- Free subscription to SberPrime+, discounts on products from partner companies
- Referral bonus for recommending friends to join the Sber team.
