Software engineering and product leadership.

Cristian is an engineer and founder who has spent 20 years building software products, teams and businesses, in sectors including e-commerce, enterprise SaaS and insurance. He has held CTO and chief product roles and still works hands-on, combining architecture and engineering with UX and product judgment. His current focus is agentic systems: coordinating AI agents that retain context across sessions and support work in engineering, operations and marketing.

• Software architecture and engineering leadership

  Architecture for web platforms and SaaS products, covering system design, third-party integrations, infrastructure and information-security processes. Cristian builds and leads engineering teams, including distributed ones, and stays hands-on.

• Product strategy and UX

  Product strategy and user experience for consumer and business software, from product direction to launch. This includes online purchase journeys in regulated insurance, where FCA rules shape what a product can say and do.

• Agentic systems and AI adoption

  Multi-agent workflows in which AI agents coordinate, keep context across sessions and work under human oversight, applied to software development, operations, customer support and marketing. Related work covers AI assistants whose access to business data follows each user’s role, and helping staff adopt AI tools in their daily work.

• E-commerce, billing and payments

  E-commerce platforms, marketplaces and online purchase journeys, with billing and payment integration. Related experience covers machine-learning personalisation, product recommendations and search engine optimisation.

• Business systems and automated decisions

  ERP and CRM systems, the integrations between them and workflow automation for repetitive processes. Automated credit-risk decisioning uses data from external providers.

• Kidfriendly.ai

  Adaptive practice for children aged 5 to 12, mapped to the National Curriculum for England. A deterministic learning engine adjusts the difficulty as a child answers, to keep short sessions challenging and manageable. Parents can see progress and what to practise next.

• Repo Prompt Overseer

  Overseer lets one AI agent supervise others in RepoPrompt CE, an open-source macOS app that helps coding agents understand a codebase before they act. A lead agent can follow parallel implementation, review and release work, send instructions and respond when a working agent asks for attention. Developers choose which sessions it oversees, and each session keeps its own approval rules.

• Minibrain

  Self-organising working memory for AI agents, kept in local files with Git history. An agent can pick up where it left off, with its goals, decisions, unfinished work and the limits on what it may do carried from one session to the next. Minibrain also keeps a list of decisions that need a person’s judgment. Coding agents such as Claude Code and Codex set it up themselves, adapting it to an existing project.

• Pleisty

  Pleisty gave online retailers machine-learning personalisation: product recommendations tuned to each page, one-to-one recommendation emails, and analytics on shopper journeys to show which audiences to target.

Cristian has written and edited software-development books for Wrox, Apress, SitePoint and Packt. Selected titles:

• Beginning ASP.NET E-Commerce in C#, Apress
• Beginning PHP and MySQL E-Commerce, Apress
• AJAX and PHP: Building Responsive Web Applications, Packt
• Professional Search Engine Optimization with PHP, Wrox