A dataholic and automation enthusiast.

Oleksii Rudenko  • Peerlist

🔹Cost saving of USD 2 million per year by integrating ML models into the pipeline and enhancing the user experience of labeling tools.

Doubled manual annotation throughput

🔹Ensure over 99% compliance of manual markup with the customers’ requirements.

Established a quality assurance process

🔹Resulting in a 15% reduction of annual data processing OPEX and 100% quality in 97% of ML data (400 terabytes of data).

Designed and implemented end-to-end data and an analytical p

🔹One of the top 20 global security market operators with an estimated capacity of EUR 15 million.

Finalized a partnership and distribution contract

🔹2-week sprints with a monthly release cadence of internal IT products, leading to ahead-of-time datasets delivery by parallelizing delivery tracks across multiple ML programs which reduced the average delivery cadence from 6 to 4 weeks.

Established an agile IT services development lifecycle

🔹More than 1400 bankers (branch sellers) trained
🔹Managing sales performed by 1300 activated branch sellers
🔹Coordination of interaction between UNIQA Life, UNIQA Group (Austria) and partner bank
🔹Sales and coaching methods and standards developed and implemented
🔹Pilot projects in Kyiv and Odessa (2011), roll-out over 32 major cities (Aug 2012), roll-out over 700+ branches all over the country (Feb 2014)
🔹Field support team developed (HC 2 in 2011, 8 in 2013, 27 since 2014)
🔹Supportive events organization (brand sales days at bank’s branches)
🔹KPI and incentive system for sellers developed, implemented and enhanced
🔹Established management accounting

Nation-wide Life insurance over-the-counter sales via the ba

🔹Handle up to 20 concurrent annotation task threads, ensuring that all new devices incorporate ML-based features by the street date.

Redesigned pipeline architecture

🔹Win a tender and secure an OEM contract with a leading European CCTV vendor with an estimated capacity of EUR 20 million.

Lead the team

🔹Completed in 6 months since the project’s inception, reducing time to throughput targets by new hires from 48 to 14 weeks by introducing pre-selection at hiring, operational excellence coaching, and expertise sharing as a part of business-as-usual operations.

Supervised the recruitment and onboarding of data processing

🔹(currently comprising 4.3 million annotated videos) from 80% to 97%+ on average by introducing human reinforced learning and automated Data Quality Control, machine-assisted and selective Data Quality Assurance.

Improved the quality of datasets

🔹 Introduced an e2e pipeline from device distribution and data collection to data annotation and delivery.