Next live sessionThu, Nov 5 · 11 am ET · 8 am PT · 16:00 UKStarts soon
Know which AI idea will pay before you spend a quarter on it.
PUSH: choose the right AI use case with maximum ROI is a free, one-hour live workshop. Six real AI ideas get scored live on impact, readiness, data, people and cost, so you can rank your own list and fund the one that reaches production. 70% of the proofs of concept we scope do.
What you'll get
- A way to rank your AI ideas by return, the same afternoon
- The test that catches a pilot that will stall, before you fund it
- A success measure your CFO will sign before any code is written
11 am ET · 8 am PT · 16:00 UK · online, one hour, free
Teams we've built for
- Comic Relief
AWS Premier Tier Services Partner, one of about 300 among more than 130,000 AWS partners.
A stalled pilot can cost a year before anyone admits it stalled.
Most pilots don't fail loudly. They stall: the data wasn't ready, nobody owned the decision, the cost per task at real volume was never worked out, and security saw it for the first time a week before launch.
Every one of those could have been asked on day one. PUSH asks them before anyone writes code, so the idea you fund is the one that can reach production, and 70% of the proofs of concept we scope do.
What you'll take away from the hour
- 01
A way to rank your list
The five scores we use in PUSH, shown on real example ideas, so your team can rank its own list the same afternoon.
- 02
The one worth funding
What separates an idea that reaches production from one that stays a demo. 70% of the proofs of concept we scope reach production.
- 03
A path to AWS funding
Which AWS programmes can fund a first proof of concept, and what they ask for.
Five dimensions, in order, because each one limits the next
PUSH puts the same five questions to every idea, in an order chosen on purpose, so an idea with a big return but no data behind it is stopped at the third question before anyone spends time on the fifth.
- 1
Business impact
The line the CFO will defend, sized on your own numbers. This is where the return is worked out, before anyone falls in love with the demo.
- 2
Technical readiness
The platforms it rides on, what's new and what's reused, and whether a managed service does the job or it has to be built.
- 3
Data availability
Does the data exist, is it clean enough, can a model reach it, and is it governed well enough to ship?
- 4
Organisational capability
Who owns it, which roles are missing, and where a partner earns a seat and where it doesn't.
- 5
Commercial model
Does it pay once it runs at production scale? We look at the cost per task, the payback and what it costs to run, because what the demo cost tells you very little.
The idea that wins is then defined the way engineers need it: the job as a change of state, never “assist” or “help”; the simplest pattern that does it (one model call, retrieval, a fixed workflow, and an AI agent only when those won't do); a named owner; what the AI agent may do alone and what needs a person's yes; and what each failure would cost.
What we bring
The idea you pick lands on parts we've already built.
Axiom
A semantic layer that gives every AI agent one meaning of revenue, customer and churn, worked out from the agreed definition instead of guessed, with each answer showing its working.
Runs in your AWS account; any AI agent reaches it over MCP.
AK AI Factory
Our method for building with AI: people write the spec and make the calls, AI agents build and test, and gates check every change before a person approves it.
Your factory, your repos, your numbers; nothing locked to us.
One hour, run by the engineers who do the work
See it
Six real AI ideas from six kinds of business, scored live on the five dimensions, and the verdict each one got: proceed, proceed with conditions, or don't.
How the winner is defined
The simplest pattern ruled out in order, the owner, the scorecard with today's baseline, and the first 20 to 40 cases with known answers.
Your questions
Bring your shortlist. We'll take as many as the hour allows and say which ones we'd look at first.
The private workshop: half a day, with the people who own the idea
We run PUSH on your own list, so every idea gets its score. The winner is defined well enough for engineers to start on it the next morning, and the first 90 days are sketched out before anyone signs anything.
Who comes
The sponsor, the person who owns the process, someone from finance, someone who knows the data, and the security lead for the last hour.
What to bring
The list of AI ideas, or the one idea everyone keeps coming back to.
Your data
Nothing has to leave your systems. We work from what your people describe, and from numbers you choose to share.
What you leave with
- Every idea scored, each with its return and a verdict
- A one-page brief for the winning use case that a team can start from on day one
- A draft scorecard: the measure, today's baseline, the target and where the number comes from
- The shape of the 90 days, with the go-or-stop at week 4
What clients say about working with us
What sets them apart is their ability to move between technical execution and business-context problem-solving.
The Armakuni team has provided both technical and product consulting on AI & automation initiatives keeping us ahead of the curve in terms of technology and optimal architecture.
The report they produced … was stand out in terms of highlighting our key risks (many of which we weren't aware of) and identifying critical next steps.
- Under 0.25%of AWS partners are Premier Tier. We're one of about 300.
- 500+AWS customer launches
- 1,300+engineers across the UK, US, UAE, India and Canada
- 70%of the proofs of concept we scope reach production
Before you sign up
Is it really free?
Yes, the hour and the half-day private version both. Scoring your ideas costs you an afternoon, nothing else.
Will you just tell us to build the biggest idea?
No. Plenty of lists end with “don't build any of these yet” and what would have to change first. Our engineers only want to start on an idea that can reach production.
We only have one idea. Is it still worth it?
Yes. Half the work is defining the one idea well: the owner, the simplest way to do it, what the AI agent may do alone and what success means in your numbers.
Who should come?
Whoever sponsors AI, whoever owns the process it would change, and someone who can speak for the numbers.
What happens after?
If an idea is worth building, our engineers start from its brief: weeks 1 to 3 agree the scorecard, there's a go-or-stop at week 4, and it's live in about 90 days.
PUSH: choose the right AI use case with maximum ROI
Thursday, November 5, 11 am ET · 8 am PT · 16:00 UK. Starts soon.
Save my seat