Ophir LOJKINE

Hyxos is a software development and consultancy company driven by a mission to implement cutting-edge digital technologies in the green energy sector. We develop a platform with the ambition to become the data backbone of all hydrogen industry players. As a consultancy, we also accompany major green electricity players in their development and their operations, focusing on leveraging the power of their data. As a CTO, my job is to oversee the tech team, transforming business ideas into working code.

• Grew the tech team from me alone to a passionate team of close-nit engineers.

• Designed a large integrated data platform with web, SCADA, and embedded components from scratch.

• Designed the mathematical optimizer making the platform smart.

• Worked at all levels, from running efficient embedded edge software in rust to running a large computing workload in Kubernetes.

We built an energy data management platform for Total that allows energy market actors to cooperate efficiently, visualize their data and extract valuable insights, and manage their devices in the field.

• Wrote Python services that now handle sensitive and complex time-series in realtime in production every day.

• I had the occasion to work both on a large kubernetes cluster and embedded real-time devices with limited resources.

• Built insightful data visualizations with Grafana.

• Modeled a large multivariate optimization problem in Tensorflow

Qwant is a search engine that does not track its users. It has the largest web index in Europe, the quality of which I contributed to. I was a member of the core 'data' team, working on the company's web crawlers, databases (indices), and data analysis tools.

• created a machine-learning model for the classification of user searches (using distributed word-embedding and clustering in Spark)

• Contributed to the crawlers, fixed many data-quality issues.

• Built data visualization tools.

Built interactive visualizations for the Roads of Russia website, a multimedia platform showcasing Russia's cultural and historical landmarks.

• Tech stack: yii, angular.js, docker, PostgreSQL

• Built some cool-looking interactive animated SVGs

ERDF (now Enedis) is the largest french electric network compagny. I interned as a data scientist, and built a workload prediction tool.

• used python with ipython, numpy, scipy and scikit-learn

• worked with an SQL database of over 10 million datapoints describing the evolution of electric consumption over ~5 years at several points in the distribution network.

• Built a web application to access the previsions.

A SQL-only low-code website builder. SQLPage is a web application framework which helps building data-centric full-stack applications by writing only SQL. It uses a set of pre-built components and a smart variable system to remove all frontend and backend programming languages from the equation, letting data people focus on what matters to them: data models and insights extraction.

Sanipasse.fr was the largest and most successful opensource EU Digital COVID Certificate verification software in France

• ~4M monthly requests served

• 100s of physical terminals, crated by multiple companies, used sanipasse throughout France

WBO is a fun side-project I'm working on in my free time. It's a free and opensource collaborative online whiteboard.

• ~4000 daily connections on the official instance, constantly growing

• Collaborating with other contributors on Github

I develop a suite of opensource tools to download zoomable images from museum websites. Also a very fun side-project, that taught me a lot.

• hundreds of websites supported

• started as a personnal tool, now has thousands of daily users

Master's degree

Courses

• Master's thesis: implement, compare, and model several distributed join algorithms on Spark

• Relational algebra

• Post-relational databases

• Queuing theory

• UML modeling

• Project management

• Human factors

Grande École (master degree)

Courses

• General engineering

• Object-oriented programming

• Databases

• Compilers

• Functional programming

an algorithm that joins relational database tables efficiently in a distributed environment using Bloom filters of an optimal size