Shamil Ulbashev

Experience

• 

  Joint Institute for High Temperatures, Russian Academy of Science

  Sept 2009 - Dec 2010

  Laboratory assistant

  • Development, assembled and start-up of a scientific experimental installation in the laboratory of extreme energy impacts.• Created and calibrated pyrometer with unique properties for experiments.

• 

  Axiomatica United Engineering Group

  Dec 2011 - Nov 2014

  Led creation of a fundamentally new device for power grid companies to reduce loses in city electric grids. Bring to the industrial prototype of innovative phase switching device for needs of power greed companies. Developed and conduce economic justification and application of energy efficiency measures for power grid companies. Proposed efficiency measures for over $1,5M anual effect.

  • Deputy Manager of the RnD department

    Apr 2013 - Nov 2014

  • Energy Auditor

    Dec 2011 - Apr 2013
• 

  SPBEC (Saint Petersburg Electrotechnical Company)

  Nov 2014 - Sept 2016

  Data Analyst/Data scientist

  • Developed virtual sensors for temperature and carbon, utilizing physical and chemical data-driven models in basic oxygen furnace. Resulted in 4% reduction in both downtime and average smelt time. Transferred approach to electric arc furnace to increase productivity and reduce costs.

• 

  LANIT

  Jun 2017 - Nov 2018

  Data analyst/ML engineer

  • Built ML and data pipelines for geo-informational analytical system for the energy sector. Constructed dashboard on BI system with advanced analytics and forecasting capabilities for data analysts at the Ministry of Energy. • Proactively initiated AI/ML venture in the metallurgy sector within the company, which led to the successful establishment of dedicated company specializing in this field that has continued operating for 5+ years

• 

  NLMK Group

  Nov 2018 - Nov 2022

  • Designed and implemented virtual sensor with advanced forecasting capabilities for sintering plant, resulting in optimized decision-making and 3% increase in prime yield.• Implemented DDMRP methodology for warehouse stock optimization, incorporating supply-and-demand forecasting through Bayesian models. Resulted in 20% reduction in stock storage usage, which in turn saves $10M per year.• Developed ML model to optimize the casting process, leading to 5% reduction in electric energy.• Independently transferred outdated locally developed ML pipelines to internal DSML cloud platform; reducing technical debt of 5 top-valued projects, increasing sustainability and reducing model retrain and deploy time from weeks to days. Guarantied shareability and composability of models.• Mentored 9 data scientists from interns to mid-levels; placing them in appropriate teams, monitoring their progress through weekly 1:1 and team meetings, and solving their big issues.• Created onboarding and orientation process for over 30 data scientists on DSML platform. Created tutorials, demos and screencasts about platform, which helped DS to perform ML engineering tasks.• Gathered feedback from DS users and worked directly with the platform development team, fostering changes that support an improved DS developer experience on the DSML platform• Assisted 8 out of 11 NLMK digital product teams with the adoption of the methodology and agile practices, cultivating a collaborative and efficient ML development environment. Show less Duties:Worked on a variety of projects to improve the steel making process and various business processes using data-driven solutions. Developed and maintained models in production sites, and conducted technical and economic evaluations.• Collaborated closely with domain experts and technical engineers to create industry solutions. Conducted thorough technical and economical evaluations that ultimately resulted in finalizing 5 successful products• Designed and created a system for optimizing the ferro-alloying process; achieving a notable 5% reduction in alloys usage for primary steel grades.• Developed and maintained model for elasticity of supply in the iron scrap market, leading to a reduction in charges by 3-5%. Convinced stakeholders to accept service by demonstrating its economic efficiency. Show less

  • Data scientist. Lead of DS methodology in center of excellence.

    Nov 2020 - Nov 2022

  • Data Scientist

    Nov 2018 - Nov 2020