The conversation
An hour on your situation, then, if it is useful, a working session with people from across your teams. We bring a rough first map and a draft list of the jobs AI has already created for you, and you correct them.
Direct the change.
Agentic AI is changing what your people do all day, and most organisations are reacting to it one tool at a time. All Present is a consulting practice that helps you direct the change.
Below are five positions on directing it, and what the work can look like: a living map of how your organisation works with AI, or a software factory your own team runs. Each is co-designed with your people and yours to take forward.
One hour with founder Jason D’Souza.
Positions
Five positions on working with agentic AI
Five things you will hear in the first hour, whatever the organisation.
Nothing compounds by accident
AI arrives through enthusiasts and contractors, each with a clever way of working that nobody else gets to use. Nothing carries from one project to the next, and pilots get built without anyone able to say why one worked. In one studio, every consultant captures a client brief their own way, so every handover rebuilds knowledge the team already paid for. The fix is a shared record of what each project learnt and an owner for every handover, so the next job starts where the last one finished.
Building got cheap. Judgement didn’t.
Agents can now build in days what took a team months, wherever the checks and the environment around them are already clear. What they cannot own is the judgement: whether something is worth building, what good looks like, and whether last month’s decision still holds. That judgement comes from experience, so give it the hours that building used to eat.
There is no AI strategy
There is an adaptable strategy for technology, or there is a document that dates as fast as the tools it mentions. A strategy that works has owners in leadership and in delivery, a short list of things they watch (what the tools can now do, what clients are asking for, what the rules require), and a set date to revisit it. It brings privacy and security review in from the first week, so a pilot never reaches the end and finds out it cannot ship. If the word AI disappeared tomorrow, a good technology strategy would still stand.
Build the engine for your work
Good technology teams start with the work: what the organisation sells, how it is delivered, where the hours go. Then they build a system for exactly that work and keep maintaining it after launch. Most AI adoption starts at the other end, with a tool, and then bends the work to fit it. Design the team for the work you want to be doing.
Presence is renewed, not defended
Somewhere in your business there is a room where a person serves a person, and that room stays human. Everywhere else, AI forces a question most organisations have not asked in years: why are we in the room, and what are we here to do? Use that question to redesign what the work is for, with the people who do it in charge. They keep their place in the room and the authority to change the work, which is what directing looks like.
Who we work with
Studios, agencies, small startups, and organisations struggling with their AI maturity
In most of them a handful of people hold the know-how, day-to-day delivery hides how far practice has drifted apart, and next year matters more than this quarter’s tools. Jason spent a decade in technical leadership at award-winning agencies, so he has sat on both sides of the table: the organisation’s and its clients’.
The canvas
A living map of how you work with AI, with the governance to back it
The canvas is a designed, visual artefact that is your living AI strategy: where AI is used and where it is not, who decides, what each team has learnt. Underneath it sits a governance baseline mapped to select published references. It also records who monitors what and what triggers a review, so the strategy gets revisited when something changes. It does not score your maturity or audit anyone’s work. Your leads review it on a schedule you agree, and in a few weeks it gets you to where a year of trial and error usually would.
The map and the baseline (fixed fee)
We fit paired interviews around your calendar, or read two or three recent pieces of work instead. You come out with the first canvas, the governance baseline with its owners, and an agreed list of the jobs. We walk you through the recommendations in conversation rather than in a report or a deck.
The knowledge system (fixed fee)
We set up one recurring ritual, a short leadership check-in, and a system in the channels you already use that feeds what each project learnt back into the canvas before the next one starts. It may include an agent, though a shared page and a habit often do the job.
The next season (scoped when you are ready)
This stage goes past efficiency to what you could build, sell or price differently now that delivery costs less. We scope it once the canvas is live.
The canvas is designed to be yours and hosted wherever suits you.
The factory
A delivery setup where your people say what to build and why, and agents build software that passes your checks
A software factory is a way of running delivery: your people write what they want built and why, and coding agents produce the software, inside gates that check the work. We design it around how your team delivers now, build it with your people and hand it over. Producers, project managers, QA and the client-side team each know their part in it and where it helps them.
- The checks are built before the agents start work, because AI speeds up whatever is already happening, good or bad.
- Rules that can be tested become automated checks. The rest get named owners and review points.
- No agent reviews its own work, and nothing ships on an agent’s say-so. Named people approve scope changes, risk exceptions and releases.
- Who did what, what was checked and what the agents were told is recorded as the work happens, ready for when a client, an auditor or a board asks.
The conversation
An hour on how your team delivers now and where the hours go.
The profile and the blueprint (fixed fee)
The profile records what your organisation needs from a factory, each line marked confirmed, assumed or unknown. The blueprint turns that into the tools, the checks, the roles and a handover plan.
The build (fixed fee)
We stand the factory up on real work, paired with your people, so the capability stays when we leave.
Your team runs it
Everything in it is yours, including the rules, the checks and what it has learnt.
We work this way ourselves: intent written down, agents doing the work, checks before anything ships.
The first factory we designed is running inside a design agency’s delivery practice.
The second is being designed for a compliance-bound software platform we help run.
About
Every seat at the table
All Present is Jason D’Souza; the bench holds multidisciplinary practitioners he has delivered with before. Jason has spent his career working alongside clients and design, delivery, development and product teams, so he knows what each seat needs from the work. The job has always been the same. Understand how a place works, build the engine for it, and leave teams capable and empowered.
Engagements run fast and light: the team is sized to the work, drawn from the bench when the work calls for it, and you will know who they are before they start.
The intro conversation
An hour on your situation
Bring where the pressure sits and what you want the organisation to look like in a year or two. We spend the hour listening, then come back with our take on your pathway and a proposal for the work.
If the booking widget does not load, email jason@allpresent.com.au.