Michael Stein

Led a 13-person data & insights org (engineering, analytics, user research) across Berlin & Helsinki, reporting to the CEO. Authored the company's data strategy and secured Board approval for implementation. Built a KPI framework, securing leadership sign-off on a 25-KPI set. Identified €5M in annual growth potential via behavioral analysis linking usage to customer value. Led a BI vendor evaluation (37 requirements, 6 themes) that avoided unnecessary tooling and migration costs.

Built the company's first behavioral & commercial analytics function for a new freemium product, defining KPIs across 4 domains and 5 dashboards for visibility. Identified and built the case for two initiatives: a lead-gen relaunch (+63% YoY leads) and a flagship product-line transformation (+56% YoY revenue). Guided business model selection and led a new product line's successful launch after two failed attempts. Delivered 9 insight studies driving decisions across acquisition, pricing, and retention.

Led a team of 3 data analysts/scientists, coordinating 50+ data models, dashboards, and insight reports. Pioneered customer-value analysis identifying 10%+ annual revenue opportunity, leading to the creation of a dedicated customer-activation team. Directed 20+ insight studies over two years (pricing, acquisition, CLV), 6 with direct executive exposure shaping strategy. Introduced Scrum, lifting productivity 30%. Grew Metabase to 60+ weekly users, 20,000+ annual dashboard views.

Developed the concepts, data models, and dashboards for user engagement, establishing Monthly Active Users as the company's first non-monetary corporate KPI. Led the migration of product analytics from Mixpanel to Metabase, building 40 dashboards and establishing it as the team's primary analytics tool. Chaired the KPI-definition process, defining 30+ metrics. Conceived "Metabase Labs," training 40+ employees in self-service analytics to scale data-informed decision-making.

Developed the Traktor Key Detection algorithm, consistently ranking among the top 3 in independent tests of commercial DJ software. Built the sound-analysis engine for sounds.com, analyzing and categorizing 2M+ loops and samples. Created novel algorithms to extract semantic information from audio as part of the EU-funded research project "Giant Steps." Core member of the Design Thinking initiative ("Inkubator") that shaped the team's approach to sound browsing and shipped the Native Browser Preview Library.

Implemented and enhanced real-time audio alignment algorithms for commercial applications: score following and drum beat tracking. Invented a directional loudspeaker and innovative sound management system based on active noise control algorithms and passive sound absorption (patented)

Worked in the Multimedia Lab / Digital Speech and Signal Processing Group on the SAPIM project ("Smart Audio Processing for Interactive Multimedia") . Developed a real-time audio alignment algorithm that improved the state of the art. Prototyped and optimized real-time algorithms for score following, drum beat tracking, percussive sound recognition, drum transcription, noise cancellation. Implemented a machine learning-based middleware package to control computer games and applications with sounds.

Developed modules of a content management system to improve the user experience of an e-learning platform Coded a voice manipulation algorithm for a Karaoke application Implemented a Tensor source separation toolbox Pursued my diploma thesis: "A method for detection and removal of audio effects in guitar and bass recordings"