Garry
Jakubiak
“Hand more of the work to agents, without losing the state or the control of it.”
I help companies identify, design, deploy and govern GenAI use cases rooted in their actual processes, through to adoption and measuring what they are worth. I also build Private Twin to explore how a reliable state of work can be kept as it moves between tools, people and agents.
role
IoD Solutions · AI consultant
The role combines building the offer and delivering it with clients:
work
Private Twin
Picture several people and several agents working in parallel on the same files. A decision is made in a meeting, an agent takes an action, a deadline moves inside a tool, a trade-off arrives by email. Yet no shared representation of the work is updated. Everyone acts on a partial version of the situation.
The problem is no longer only finding information. It is knowing what is true right now, what has been done, what is still expected, who carries the next step, and what an agent may commit to without a fresh human check.
Private Twin explores the infrastructure this new organisation of work needs: a current state that is sourced, correctable and shared between people and agents. It brings expected work close to the signals of observed work, to produce a defensible timesheet, a daily readout and the next actions.
Explore Private Twin
A working day rebuilt from authorised traces, with the time attributed to the matching projects.
DD Analysis
DD Analysis simulated a multi-echelon supply chain and compared inventory policies in an environment exposed to demand and supply variability.
I wrote the algorithms in C#, built the offer and sold the solution to paying clients for eighteen months. The project also included an advanced clustering approach, segmenting product families beyond ABC/XYZ classification. This earlier work shows how I turn domain expertise into a product, then take it to market.
Explore DD Analysis
Inspect 
Inspect One consistent way of working
In client engagements as in my own projects, I usually work across several steps of the same chain:
Understand
Analyse the problem, the ground reality and the constraints.
Build
Turn the findings into a method, an offer or a solution.
Set in motion
Present the approach, bring the teams in and organise delivery.
Confront
Test, measure, deploy and support adoption.
Notes from the field and from the build
A place to write about GenAI transformation, the move from reactive agents to more autonomous ones, and the state of work. Each piece states plainly what was observed, what was tested and what is still a hypothesis.
Read the writingMore than twelve years across supply chain, transformation, decision systems and artificial intelligence.
See the full backgroundGet in touch
To talk about a GenAI transformation, agents, Private Twin or my background: