A strong fit
- Professional firms of 10–100 people: architecture, engineering, legal, real estate
- Developers and operators with repetitive document work
- Founders who want an AI operation that belongs to the business
02
SERVICES · SYSTEMS
AI and operations systems.
Less manual work in planning, building, selling and reporting.
THE APPROACH
AI that pays back, in the right order.We help professional firms and operators decide where AI starts, prove it on one measured problem, and hand over a system the team owns.
WAYS TO START
Start small. Each step stands on its own.
Free to learn, a paid hour to decide, a fixed-price sprint to prove it.
- 01
Free webinar, in Hebrew
Five real cases from real-estate projects, and one audience problem worked on screen.
FreeTue 20 Oct · 90 min · Google Meet - 02
AI Strategy Session
An hour on your firm: where AI should start, what you already pay for and what to measure. The fee is credited toward the Audit Sprint.
€15060 min · Google Meet - 03
Audit Sprint
Stage 0, focused discovery, as a fixed-price sprint: a decisions document, a tiered roadmap and agreed success measures. Limited slots.
Fixed price, by proposalTwo to three sessions - 04
Guided pilot and tapering support
A pilot on one work domain, then support that tapers off and ends at month twelve.
By proposal, after the sprint3–5 weeks, then up to 12 months - 05
RealizeOS setup assistance
An hour to install and adapt RealizeOS, our self-hosted AI operations system.
60 min · online
WHO IT’S FOR
Busy teams with the tools, but not the time.
The gap is rarely access or understanding. It is deciding what to do first, who owns it and protecting the hours to do it.
Not yet a fit
- Teams looking for a tool demo, not a change in how work gets done
- Programmes with no one who can give four hours a week
- Projects that must start with the most sensitive data
THE ENGAGEMENT
Separate stages. A stop point after each.
There is no commitment to the sequence. The retainer tapers on a schedule written into the proposal.
Focused discovery
Two to three in-depth sessions. We name the problem in your words, take three management decisions and check what you already pay for.
You get: A decisions document, a tiered roadmap and agreed success measures — yours to keep.
Guided pilot
Three to five weeks, one work domain, two people. We choose the tools and guide the build; your team operates and tests it.
You get: A working process in production, measured before and after on real cases.
Tapering support
Around ten to twelve advisory hours a month, tapering from full to half to a quarter and ending at month twelve.
You get: A team that runs and extends the system without us.
Team workshops
Sessions built on your own cases: how to spot where AI fits and match the right tool to it.
You get: People who find the next use case themselves.
GROUND RULES
Four principles, agreed before any tool.
Tools change. These are what keep a programme alive in a busy organisation.
Buy first, build for the gap
We test what the market offers and what you already license before anyone writes code.
Human in the loop, always
The system prepares the work. Professional judgement stays with your people.
The knowledge stays with you
Everything is documented and handed over. Our support is designed to end.
Start small, measure honestly
One domain, a small group, a pass mark agreed in advance — and a stop rule if it misses.
PROOF
The method, in a real firm.
A 25-person architecture practice: eight domains mapped, one pilot chosen, break-even modelled at month seven.
ADVISORY CASE STUDY · PROFESSIONAL SERVICES
Case study
AI adoption roadmap
How one engagement moved a 25-person firm from scattered experiments to one measured programme.
View the case — AI adoption roadmapSYSTEMS WE BUILT AND RUN
We use what we recommend.
Our own operations run on AI systems we designed. They are options for you, never a requirement.
AI OPERATING SYSTEM · OPEN CORELive
RealizeOS
A self-hosted team of AI agents that knows the business, remembers and works under human approval.
View the case — RealizeOSMEETING INTELLIGENCE · HEBREW-FIRST
Live
MeetSum
Every meeting becomes searchable memory, tasks and automation — in Hebrew and English.
View the case — MeetSum
LIFE-DESIGN APP · PRIVATE BY DESIGNLaunching
Dreamward
A private app that turns the life you envision into the life you live, with an AI companion on your own AI.
View the case — DreamwardALSO AVAILABLE
Fractional operations and development management.
Asaf Eyzenkot takes a limited number of B2B roles: operating models, project coordination and real-estate development management.
IN THEIR WORDS
What clients and partners say.
Asaf has developed strong expertise in helping design and build AI-driven systems that support smarter decision-making, efficiency, and scalable business structures.
WATCH
AI in real projects, in short videos.
FIELD NOTES
AI in practice, from real projects.
Short notes on what worked, what did not and what it cost.
AI in practice
Five hours to build. A month to trust.
A tender agent is quick to build. Earning the office's trust takes longer.
AI in practice
How to run an AI pilot a firm can actually judge
One domain, two people, three to five weeks, and a pass mark agreed before anything is built.
AI in practice
Why AI adoption fails at step three
Firms start with the most valuable problem. They should start with the most visible one.
AI in practice
Making Hebrew speech recognition production-safe
A 52-minute mixed-language meeting crashed our local model. The fix was a routing policy, not a bigger server.
QUESTIONS
Before you ask.
How quickly will we see a result?
Discovery takes two to three sessions. A pilot runs three to five weeks, and most of that time is quality checks against real cases rather than building. In our modelled case, the programme breaks even in month seven.
How much of our team’s time does it take?
Two kinds of time. One internal lead for four to six hours a week, and two to three hours a week from each pilot participant, mostly reviewing outputs. We ask for those hours in the proposal, because without them the work does not happen.
Do we have to use RealizeOS or your other systems?
No. We start with what you already pay for and what the market offers. RealizeOS, MeetSum and our other systems are options when they fit the gap, not a requirement.
What does it cost?
Each stage is priced as a separate, fixed unit before it starts, and you can stop after any stage. Discovery is deliberately a small first commitment. Ask for a proposal after a 20-minute intro call.
Where do you work, and in which languages?
Remotely, and in person in the Lisbon area and Barcelona. We work in English and Hebrew, with basic Portuguese and Spanish.
Can you also run operations or a development project for us?
Yes. Asaf Eyzenkot takes a limited number of fractional operations and development-management roles, contracted through Realization Unipessoal LDA.
START WITH DISCOVERY
Where should AI start in your firm?
Twenty minutes is enough to know whether discovery is worth it.