Kliment Merzlyakov, Developer in Mexico City, Mexico
Kliment is available for hire
Hire Kliment

Kliment Merzlyakov

Verified Expert  in Engineering

Bio

Kliment is a business-oriented data scientist with several years of industry experience. With an aspiration to make machine learning projects done without quality trade-offs, Kliment can handle any project, from a business hypothesis brainstorming session to implementing a model and final integration.

Portfolio

Plytrix Analytics
BigQuery, Snowflake, Data Build Tool (dbt), dbt Cloud, Fivetran, Python, SQL...
Vigo
Data Science, Machine Learning, ClickHouse, Cassandra, Spark, Scala...
Airpush
Neural Networks, Web Scraping, Scikit-learn, Machine Learning, Data Science...

Experience

  • SQL - 6 years
  • Python - 6 years
  • Data Science - 6 years
  • Machine Learning - 6 years
  • Web Scraping - 4 years
  • Generative Pre-trained Transformers (GPT) - 3 years
  • ClickHouse - 3 years
  • Natural Language Processing (NLP) - 3 years

Availability

Part-time

Preferred Environment

MacOS, PyCharm, Python

The most amazing...

...machine learning result was the implementation of my preprocessing algorithm in a recommender system in the largest Russian eCommerce retailer, doubling CTR.

Work Experience

Senior Analytics Engineer

2022 - PRESENT
Plytrix Analytics
  • Led a three-engineer team to develop a robust data integration solution. Designed and implemented processes for ingesting data from 100+ sources. Orchestrated the creation of 300+ dbt models and 700+ data validation tests.
  • Decreased the average win-back communication period from 66 to 42 days by integrating a machine learning churn and LTV model using Google Cloud Functions.
  • Reduced miscommunication between business units by creating unified LookML and dbt layers. Implemented Looker Explores with GCP, Fivetran, dbt, and Snowflake. Shrank data discrepancy from 4% to 0.3%.
Technologies: BigQuery, Snowflake, Data Build Tool (dbt), dbt Cloud, Fivetran, Python, SQL, ETL, Databricks, Google Cloud Platform (GCP), Looker, Microsoft Power BI, Superset, Churn Analysis, Consumer Behavior, Data Analytics (Marketing), Marketing Analytics, Data Engineering

Senior Data Scientist

2020 - PRESENT
Vigo
  • Developed and integrated machine learning model to predict customer satisfaction index, which telecom operators use to optimize infrastructure expenses. Technologies used: Spark with Scala, Apache Airflow, CatBoost, Cassandra.
  • Developed and integrated machine learning model to clean up the data. There are a hundred thousand events per second sent from users and sometimes this data is wrong—the developed model filters out non-relevant data with 0.97 ROC-AUC.
  • Integrated data science technologies into the company's development pipeline: Apache Airflow, ClickHouse, Redash.
Technologies: Data Science, Machine Learning, ClickHouse, Cassandra, Spark, Scala, Apache Airflow, SQL, CatBoost

Data Scientist

2017 - 2020
Airpush
  • Developed and delivered an adult detector for apps and ad creatives which helped to keep the ad network brand safe.
  • Developed and integrated a CTR predictor, which increased revenue by 2%.
  • Integrated several Slackbots with analytics and notifications.
  • Integrated business intelligence tools Redash and Looker.
Technologies: Neural Networks, Web Scraping, Scikit-learn, Machine Learning, Data Science, SQL, Redash, Looker, ClickHouse, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), Python

Data Scientist

2013 - 2017
Ulmart
  • Created an algorithm and also integrated it into the company’s recommender system, which increased CTR by 200%.
  • Developed and maintained ab analytical system for a net promoter score project, which was used by the KPI system.
  • Applied an LTV analysis for marketing expenses optimization.
  • Integrated Tableau and provided stakeholders with dashboards.
Technologies: Web Scraping, Scikit-learn, Machine Learning, Data Science, SQL, Tableau, BigQuery, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), R, Python

Experience

Adult Detector at Airpush

This is an ensemble of models to predict adult content based on picture, description, and name. I collaborated with a team of Android developers to deliver the model to production. My part was to define and collect the needed data, develop a model, and integrate a Flask service. The model was successfully implemented, and the back end could obtain an adult score by request.

Education

2017 - 2019

Master's Degree in Mathematics and Applied Statistics

Saint Petersburg State University - Saint Petersburg, Russia

2009 - 2014

Bachelor's Degree in Management

Saint Petersburg State University - Saint Petersburg, Russia

Certifications

AUGUST 2022 - AUGUST 2025

AWS Certified Cloud Practitioner

Amazon Web Services

DECEMBER 2020 - DECEMBER 2021

Certified LookML Developer

Looker

Skills

Libraries/APIs

Scikit-learn, PyTorch, CatBoost

Tools

Looker, Redash, PyCharm, BigQuery, Tableau, Apache Airflow, dbt Cloud, Microsoft Power BI, Superset

Languages

SQL, Python, R, Scala, Snowflake

Storage

ClickHouse, Cassandra

Frameworks

Spark

Paradigms

ETL

Platforms

Amazon Web Services (AWS), Databricks, Google Cloud Platform (GCP)

Other

Data Science, Machine Learning, Web Scraping, Natural Language Processing (NLP), Neural Networks, Economics, Statistics, Generative Pre-trained Transformers (GPT), Data Build Tool (dbt), Fivetran, Churn Analysis, Consumer Behavior, Data Analytics (Marketing), Marketing Analytics, Data Engineering

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring