Nedvision.ai is a Skolkovo resident, MTK. Salary range - from 100 thousand rubles.

Real Estate Investment Scoring Format: contract or part-time We are looking for a promising junior / junior+ specialist who wants to grow in the field of Applied ML Engineering: to work with real dirty data, build features and models, bring solutions to practical use, and gradually take on more responsibility.

We don't need someone who already knows everything. We need someone who loves working with data, knows the basics well, learns quickly, is not afraid of uncertainty, and wants to develop on a live product task in the field of real estate investment scoring. What you will be doing: Work with the source data on real estate: listings, texts, prices, geodata, seller behavior, market dynamics.

Perform EDA: look for patterns, test hypotheses, find problems in the data. Prepare features for ML models: text, tabular, time-series, geographical. Assist in building models for evaluating the investment attractiveness of properties: rating, fair price, profitability, liquidity.

Participate in model quality assessment: selecting metrics, validation, error analysis, interpreting results. Analyze quality degradation by segments: districts, property types, price categories, periods. Gradually take on more complex tasks: from analysis and prototypes to independent solutions for individual parts of the ML pipeline. Help structure the ML process: describe hypotheses, record experiment results, maintain reproducibility. What you need to know at the start: Python at a level of confident data handling: pandas, numpy, notebooks/scripts. Basic ML: classification, regression, train/test split, quality metrics, overfitting, leakage. Basic statistics: distributions, correlations, hypothesis testing, confidence intervals. SQL: selections, join, group by, window functions are a plus. Understanding of feature engineering: how to turn raw data into features. Ability to carefully work with dirty, incomplete, and noisy data. Ability to explain your findings in simple terms: what was tested, what was the result, why it is important. Willingness to delve into the subject area, ask questions, and bring tasks to completion.

Will be a plus

• Experience with scikit-learn, CatBoost / LightGBM / XGBoost.
• Basic understanding of NLP: embeddings, sentence-transformers, features from texts.
• Interest in geoanalytics: coordinates, distances, districts, POI, H3, PostGIS / GeoPandas.
• Experience with Airflow / MLflow / experiment tracking.
• Experience with marketplace / classifieds / real estate / fintech data.
• Ability to use AI coding tools to speed up work without loss of quality.

What we will teach you

• Working with geoanalytics in applied ML tasks.
• Building features for real estate objects: location, transport accessibility, surroundings, district, traffic, market dynamics.
• Understanding the real estate domain: price, liquidity, profitability, listing quality, seller behavior.
• Honestly assessing ML models: time-based split, backtesting, out-of-time validation, analysis of degradation by segments.
• Bringing ML solutions from hypothesis and prototype to practical use in the product.
• Working with tasks where there is no ready-made educational answer, but you need to think, test, and make informed decisions.

What we will provide

• Plenty of practice with real data and real business tasks.
• Opportunity for rapid growth in Applied ML Engineering.
• Access to a strong subject task at the intersection of ML, real estate, geoanalytics, and investment scoring.
• Carte blanche for justified decisions: if a hypothesis, approach, or tool is well-argued, it can be tried.
• Opportunity to take on not only technical but also semantic and organizational tasks: formulate hypotheses, structure experiments, suggest process improvements.
• Work in a contract or part-time format with clear tasks and gradual growth of responsibility.

**Resume: **recruiter contact