Profileimage by Martin Tabikh Machine Learning Engineer from Paris

Martin Tabikh


Last update: 21.09.2022

Machine Learning Engineer

Company: Martin Tabikh
Graduation: PhD Computer Science
Hourly-/Daily rates: show
negotiable depending on the project
Languages: English (Full Professional) | French (Native or Bilingual)




AWS, Analytics, Programming, Python, C++, Pytest, object-oriented coding, Unix, Docker, math, time series, Pytorch, RNN, Pandas, Numpy, Gensim, Scikit-learn, SpaCy, Matlab, Databases, Code versioning, Cloud, Snowflake, Postgres, git, Gerrit, Github, Azure, Azure devops, Machine Learning, databases, Azure ML, Python libraries, algorithms, data science, APIs, user stories, coding, wireless networks

Project history

01/2022 - 10/2022
Machine Learning Engineer
Pernod Ricard
* Implemented CI/CD pipelines with Azure, managed releases and generated release notes automatically.
* Developed a Streamlit web app and an api with multiple endpoints triggered daily.
* Managed the config and connections with different Snowflake databases.
* Optimized code and decreased execution time by 80%.
* Created Docker files for dev/test environments.
* Proposed, monitored and published drift metrics using Azure ML.

02/2021 - 01/2022
Machine Learning Engineer - Data Scientist
* Supported researchers/data scientists in the team to improve their code and make it more reproducible.
* Developed Python libraries for generative classification models.
* Developed hyper-parameter optimization algorithms such as successive halving and Bayesian hyperopt.
* Developed several functionalities for the Machine Learning pipelines creation tool used internally.

10/2019 - 01/2021
Machine Learning Engineer
* Built machine learning features to support code bug categorization and retrieve relevant developers.
* Trained (un) supervised models, designed APIs and apps in a team of 8 engineers and data scientists.
* Decreased time spent to correct bug issues by 50% leading to fewer expenses (~2Million euros per month)
and increased customer satisfaction consolidating Nokia's position in the market.

03/2018 - 09/2019
Software Engineer
5G: Nokia
* Implemented 2 top-priority features increasing 5G's spectral efficiency and bandwidth.
* Prepared user stories, wrote, tested, and reviewed code in a team of 8 software engineers.
* Created an environment of productivity that was instrumental in reducing coding time by 20%.
* Achieved ~1Gbps (x10 compared to 4G), the highest throughput for wireless networks.

03/2015 - 03/2018
Software Engineer
Orange is France's number one telecom operator and a company that specializes in cybersecurity risk aversion, cloud computing, and
connectivity solutions and is represented throughout Europe and Africa.

* Proposed new 5G algorithms optimizing the beamforming procedure (coding information streams to
cancel interference between different mobiles).
* Collaborated with 5 scientists to mathematically model 5G beamforming, then coded models to enrich
Orange's 5G simulators.
* Introduced findings to international conferences and telecom consortiums.
* Filed a patent to protect the originality of the algorithms and to increase Orange's revenue.

03/2015 - 09/2015
Data Scientist
Agoranov is a mid-sized (~50 people) incubator dedicated to the development and promotion of innovative entrepreneurs focusing
on industry, green tech, and health.

* Introduced a statistical model explaining 10 success factors of startups.
* Analyzed data from CrunchBase dataset and other internal sources in a team of 3 analysts, using R.
* Published the study internally to the benefit of all incubated startups.

02/2014 - 09/2014
Project Manager
* Analyzed cost-effective solutions to reduce latency in 3G/4G networks.
* Led the testing phase, contributed to choosing solutions and coordinated Proof of Concepts. Worked in a
team of 6 people from diverse backgrounds.
* Helped decrease Orange's 3G and 4G latency by 30%, increased customer's quality of experience, and
decreased retention by at least 10%.

Time and spatial flexibility

Remote/Paris/Could travel occasionally

Contact form

Contact details