AI training

Build the capability.Keep the code.

Hands-on AI workshops in your own stack. Your engineers build an agent, test how it fails and harden it, guided by engineers who ship AI in production.

AWS Advanced Tier Services Partner · OpenAI Partner · 5 certified engineers
What you get

Your engineers build it. Your team keeps it.

Your account. Your repositories. Scope agreed before work starts.

01 /
A working repository

The code and exercises your team creates.

02 /
Practical security skills

Permission boundaries and failure-path testing.

03 /
A path to apply it

Learning materials and next steps for your team.

An AI workshop learning path
Illustrative learning path; the workshop scope is agreed with your team.
Our work

Training on real work, not toy demos.

View all work

We also build and operate software ourselves. tMinus1 built Automatum, a multi-tenant platform with marketplace integrations and metered billing. That platform work informs how we approach identity, integration, deployment and operational ownership.

Platform engineering experience. The case study is not a claimed result from this service.

Automatum case study
The service in practice

Build capability your team can use.

Hands-on AI workshops for engineers, security teams and leaders. Work through the implementation, the failure paths and the decisions around a real workflow, then keep the code and materials for the next project.

BEFORE

Scope the cohort and exercise

Discuss your stack, participants and objectives with the engineering lead. Agree prerequisites, environment setup, data and the process the group will work through.

DAY 01

Foundations through a working build

Introduce the architecture and failure modes in context. Build an initial workflow and inspect the requests, responses and connected components together.

DAY 02

Integration and evaluation

Develop the workflow, connect agreed interfaces and build test cases. Review permissions, usage and error handling alongside the application code.

DAY 03

Challenge, harden and explain

Run controlled adversarial exercises and refine the implementation. Walk through the evidence and agree the next steps. Optional deployment work is scoped separately and follows your change process.

Scope & capability

Learn the whole workflow.

The engineering workshop follows the work from first request to operating responsibility. Your team sees where the model fits, how it connects to other systems and what needs to be tested before anyone depends on the result.

FOUNDATIONS

Understand the moving parts.

Work through model requests, context, retrieval, orchestration and tool calls in an application your engineers can inspect. Discuss where conventional software is a better fit and how to recognise a workflow that is too broad for a first build.

INTEGRATION

Connect a bounded task.

Build against agreed interfaces in a non-production environment. Define inputs, outputs and the permissions a tool needs. Work through validation, error handling and the point at which the system should ask a person to decide.

EVALUATION

Make quality testable.

Create representative examples and difficult cases, then inspect the behaviour. Review answers, tool use and failure handling against agreed expectations. Use the exercises to understand how changes to prompts, context or configuration affect the workflow.

SECURITY

Challenge what you built.

Exercise prompt injection, permission boundaries and data exposure scenarios within the workshop environment. Investigate the observed behaviour, refine controls and document the limitations. The goal is a repeatable engineering approach your team can apply again.

Illustrative workshop learning path

Build it. Inspect it. Improve it.

A useful engagement connects the evidence to a practical action. This example shows how we structure that reasoning; actual findings and deliverables depend on the agreed scope.

BUILD

A working task

Choose a bounded exercise and connect its application, context and tools.

OBSERVE

Evidence of behaviour

Collect representative outputs, tool traces and failure cases for review.

CHALLENGE

Agreed adversarial exercises

Try to redirect the workflow, cross permission boundaries or expose data in a controlled lab.

HARDEN

A revised implementation

Apply controls, rerun the cases and explain what changed and what remains uncertain.

Illustrative structure, not customer findings or a completed engagement.

What you take away

Deliverables your team can work with.

The value of the engagement continues after the walkthrough. Your team keeps the agreed implementation or materials, the findings and the context needed to decide what happens next.

CODE

A repository to return to

Keep the workshop implementation, exercises and configuration agreed for the cohort. Your engineers can revisit the decisions and use the work as a reference for the next internal project.

TESTS

Evaluation and security exercises

Retain the representative cases and lab scenarios used during the workshop. These provide a starting point for evaluating later changes rather than relying on the original demonstration.

CONTEXT

Design and operating notes

Document the workflow boundaries, permission choices, known limitations and the operating considerations covered in the sessions. The materials connect implementation decisions to their purpose.

NEXT STEPS

A practical continuation plan

Discuss what needs to happen before applying the learning to a live project. Separate workshop outcomes from the access, assurance and release work needed for production.

The important detail

Different roles need different depth.

The engineering track is the hands-on core. Security and leadership sessions focus on the decisions those teams own. We agree the audience and agenda in advance so participants can work at the right level.

For engineers · three-day core

Build a scoped workflow, connect tools, create evaluation cases and harden the boundaries. Participants should be comfortable writing and shipping code in your stack; a machine learning background is not required.

For security teams · one-day track

Work through agent identity, tool permissions, data flows and review evidence. Controlled demonstrations make the failure modes concrete, then the group develops a checklist relevant to its own review responsibilities.

For leaders · half-day track

Discuss how to choose a first workflow, define ownership and assess the evidence behind a pilot. Work through funding decisions, operating responsibilities and the conditions needed before expanding the use case.

Preparing for the engagement

Set the room up for useful work.

A prepared environment and a shared exercise let the workshop spend time on engineering. We confirm the setup with your lead rather than asking the cohort to improvise access on the first morning.

A safe learning environment

Use a non-production environment and agreed synthetic or anonymised data. Confirm accounts, repositories and permissions before the build sessions. Production access is not required for the core workshop.

The right participants

Keep the build sessions focused on engineers who can work with the selected stack. Invite security colleagues to the relevant demonstrations and arrange leadership discussions separately where appropriate.

A clear agreement

The agenda, cohort, follow-up sessions and any optional deployment work are agreed in scope. Pricing is quoted per cohort after the scoping conversation; travel is agreed separately where needed.

Useful answers

Before we start.

Discuss anything specific to your team or environment.

Is this official AWS training?

No, and we will not imply otherwise. We are an AWS Advanced Tier Services Partner, not an accredited training provider, so there is no official curriculum and no AWS certificate at the end. If you need accredited courses, we will point you at them.

What do our engineers need to know already?

They should be able to write and ship code in your stack. No machine learning background is needed. If your team can build and deploy a service, they can do this week.

Do you need access to our production systems?

Not for the three-day core. We teach against a non-production environment and anonymised or synthetic data where that is what you prefer. The optional fourth day is the only part that touches a real environment, and only through your change process.

Can we mix engineers, security and leadership in one group?

Keep the build days to engineers. The attack demonstration on day 3 is worth opening to your security team, and the leadership session works better separately and shorter.

What does it cost?

One fee per cohort, quoted after the scoping call, because the agenda depends on your stack and the process you pick. Travel is quoted separately and at cost.

The next step

Train the team that will run it.

Send a little context. We’ll reply to arrange a technical conversation; submitting the form does not book a time.

Request a technical call
AI training — your stack, your codebase, code your team keeps
Request a technical call