Skip to content

AI training for teams

Tokyo teaches your company on how to properly use AI. In the end, teams learn how to coordinate their entire workflow using AI Agents.

Your team already uses AI, each on their own. What is missing is knowing what to ask it for together.

Co-founder of Geometric Fever, CTO of Plural and Humanform.

Ask for a training

The builder

There are companies working with his AI and engineers building on the code he wrote.

In companies

  • At Plural, a jobs platform, he has been CTO since 2022.

    • Search: he built an AI search that understands natural language. Clicks from search rose 30%, and hires 20%.
    • Data: an AI recommendation system, correctable by a person, removed more than 99% of invalid job titles. And AI agents cleaned and enriched a database.
    • Agents: he created evaluations that cut hallucinations from AI agents by 90%.

  • At Humanform, he is CTO. The company creates real-time face and voice clones that belong to each user.

  • At Geometric Fever, he built Pascal, the AI agent that works inside the company.

    • The site you are reading was built by AI agents, under rules he wrote and published.
    Meet Pascal

In open source

  • The open-source libraries he published have been downloaded more than 23 million times.
  • There are 1682 public repositories and 26 packages that depend on the most downloaded one.
  • He wrote instructions that AI agents load in order to work with rigour. They have been installed 930 times.

Each set has a published test: the same model, with and without them, with the negative results included.

The teacher

Tokyo takes complex technology and gets people using it.

A team he trained in AI rated him on what he knows about AI and on the way he teaches. The average of the answers was 9.2 out of 10.

One of the things it valued most was the clarity: complex technical concepts, broken into practical ideas that do not intimidate.

  • With managers: he was the CEO’s technical right hand. He translated the vision into an engineering plan and prepared the reports for the board and for investors.

  • With engineers: he mentored seven of them into full-stack developers. He cut onboarding for new arrivals from 8 days to 2, with code in production on the first day.

    At Plural, he moved the engineering team onto AI tools: it got 20% faster, at the same quality.

  • With students: as a teaching assistant at the German University in Cairo, he taught algorithms and programming. In one semester, student satisfaction rose from 40% to 85%.

In the room

Hands on and real cases. At the end, the team knows what it wants to ask the AI for.

  • Practical exercises: building requests to the AI that turn into business processes.

  • Real business cases: AI working inside a company, shown step by step.

At the end, that team asked for a second session: automation and workflows in each department.