About

Who actually does the work

Ittai Perez

Founder, Dynamic Reactions LLC. Faro, Portugal.

Most AI programs stop where the work begins. People leave a workshop able to describe a tool, then return to a process with no room for it. Adoption is a design problem, not an enthusiasm problem, and I spent twenty years on that side of it: what people practise, in what order, to what standard of done.

Now I also build the system that holds the work. StartupKid runs its mentor recruitment platform in production because I saw the bottleneck from inside the CLO role and built the tool. A school’s teaching staff trained on a workshop app I built in a day for their session. The deck is not the deliverable. The running system is.

Dynamic Reactions is my practice. I lead every engagement myself, with the two specialists below.

Background

StartupKid

Chief Learning Officer. I own the pedagogy and content architecture behind the programs, and the mentor side of the work: how mentors are recruited, trained and developed, and the platform that recruits them.

Teach Beyond Borders

Chief Learning Officer. I designed and delivered training programs for Israeli municipalities, manufacturers and companies, from the brief to the room.

eTeacher Group

Head of Training and Quality for eight years. I built and ran the training organization of an online education group: more than 150 instructors across 15 countries, their recruitment, training and quality standard, and Tekkie Uni, the online STEAM school I built there.

Speaking

Talks on learning and AI for organizations in Israel and abroad, including the Israeli AI communities, and training sessions for teaching staff and consulting teams.

LinkedIn

The team

I work as a lead with two specialists, not as a firm with a bench. I own the learning design and the build; they own what makes the result safe to run and connected to everything else.

• 

  Bar Abudi

  Production and governance

  Bar runs production. He has managed environments serving hundreds of users, Kubernetes-based AI platforms among them, across cloud and on-premise infrastructure: CI/CD, Git workflows, internal APIs, automation, and local and hosted AI models. He is what turns a fast prototype into something an IT department will sign off on.

  For AI-assisted development he sets the working rules: version control, testing, documentation, code review, deployment, secrets and access management, and how much technical review each change needs. He has taught programming for years, and it shows: he brings teams into structured practice from whatever level they start.

  LinkedIn
• 

  Shira Segev

  Automation and integration

  Shira works on the seam between logical design and technical execution. Her ground is language models, integration and data, architecture and product management. She approaches a project as a system first, modelling it in UML and BPMN to draw the boundaries and get the logic right, and then builds it: Python and SQL, API integrations through tools such as Make and ManyChat, and security testing (DAST and SAST) so the architecture stays stable under use.

  With Claude Code she builds projects, sites and software, pairing AI-assisted work with code she owns. In an engagement she is the connective tissue: the automations, the integrations and the specs that keep a build honest.

  LinkedIn