00/About
Alex
Product leader and engineer by training. 18+ years in IT, 14 in management, teams of 80+.
00-A/What I’ve owned
No CV. Here’s what I’ve been accountable for.
- Domains
- Enterprise data and master data (MDM) · ERP · Business process management (BPM) · Industrial procurement · Digitalization across production, maintenance, safety and logistics · Applied ML and AI agents
- Industries
- Steel · Rail freight logistics · Banking · Car-care services (as startup CEO) · Earlier, as a systems integrator: telecom, energy, the public sector and aerospace engineering
- ML in production since 2016
- Banking: graph analytics of links between corporate borrowers to manage credit risk · Rail freight: an ML forecast of freight demand across an entire national network, feeding the sales plan · Steel: an AI substitute finder and catalog deduplication
- Roles
- Startup CEO · Corporate CPO · Before that, IT analyst at Boeing: I did the analysis myself long before I commissioned it from anyone else.
- Scope
- Full P&L as a startup CEO · Corporate portfolio budgets with a 5-year IRR I answer for: costs against hard savings · Investment plans and portfolio business models defended before top management · Technology choices defended at IT committees · Portfolios of enterprise systems · Cross-functional teams of analysts, engineers, data and design specialists · Hiring, and parting ways when it’s needed
- How I run teams
- In 2022 I began phasing out standalone project managers. Today my teams run without them: small groups of T-shaped people own the result end to end. In my own products, AI agents do the build.
- Credentials
- Engineering degree · Master’s in project management, taught in English · Corporate mini-MBA · PRINCE2 · English: C2
00-B/Bio
Bio
Alexander (Alex) Sokolov is a product executive and engineer by training. In 18+ years in IT he went from IT analyst at Boeing, supporting CAD/CAM/CAE systems, to startup CEO and corporate CPO. He spent 14 of those years in management, and his largest cross-functional teams numbered 80+. As a startup CEO he ran the full P&L. In corporate roles he has owned enterprise data and master data, ERP and BPM portfolios, industrial procurement digitalization and applied ML, and answered for each portfolio’s 5-year IRR: every cost, people first, weighed against hard savings in live operations.
His method is simple: data first, expert second, IT third. Every initiative gets a KPI that shows up in the P&L, an owner and a deadline. Sometimes the honest advice is not to use AI at all.
He builds AI products with AI agents as his engineering team. He designs, decides and directs; the agents write the code behind automated test gates. His work includes an LLM agent that finds substitute parts in inventory data; Aristarkh Gradov, an open-source AI persona engine; Evolution Workbench, a physics game; and The Rat Box, a simulation built on 15 social-science models.
Online he is alexprinceps, a nod to the Roman princeps senatus, first among equals.