Train the team with the technology we deliver

1 day - Management, team leads, office staff

Generative AI & LLMs for Your Company

What private language models can write, summarise and translate in your company's own words, and how to choose the first process to automate.

• Private models vs public cloud
• Drafting, summarising, translating
• Verifier and human approval
• Choosing the first use case

Technology: Generative AI & LLMs →

1 day - Team leads, process owners, office staff

AI Agents & Copilots at Work

How agents carry out multi-step tasks in your CRM, ERP and ticketing systems, and how copilots assist each employee in the tools they already use.

• Agent vs copilot
• Rules and permissions
• Hand-over to a person
• Audit trail of every action

Technology: AI Agents & Copilots →

1 day, hands-on - Knowledge owners, IT, office staff

RAG & Knowledge Search on Your Documents

Chat with your own manuals, contracts and procedures: answers built from your documents, each one with a link to its source.

• Preparing your documents
• Answers with cited sources
• Access rights
• Checking answer quality

Technology: RAG & Knowledge Search →

2 days, hands-on - Software, QA and DevOps engineers

Agentic Pipelines & Verifiers

Build an agentic workflow inside Jira, GitHub and CI with a measured verifier and human accept/reject on every step.

• Agent framework and integrations
• Verifier design and mutation testing
• From agentic to deterministic scripts
• AI Security Gateway

Technology: Agentic AI Data Pipeline →

1 day, hands-on - Software and IT engineers

Agents in Teams, Slack, Email and Ticketing

Connect agents to the channels your teams already use, so requests come in and approved answers go back the same way.

• Channel connectors
• Draft, approve, send
• Tokenising sensitive fields
• Adding a new channel

Technology: Communication Tools Integration →

1 day, hands-on - ML and software engineers

Voice Recognition for Your Vocabulary

Adapt a small speech model to your product names, procedure steps and accents, and use it for dictation and hands-free commands.

• Preparing recordings
• Fine-tuning on your terms and accents
• Dictation and operator commands
• On-premise and on-device processing

Technology: Voice Recognition →

1 day - Marketing, HR, customer service

AI Avatars & Digital Humans

Set up a talking avatar for reception, sales, training or a first interview, answering from your own knowledge base.

• Voice, language model and face
• Answers from your knowledge base
• Hand-over to a person
• Consent for faces and voices

Technology: AI Avatars & Digital Humans →

1 day - Engineers, product owners, operations leads

Physical AI Foundations

What small vision and decision models can and cannot do on the floor, and how to scope a first MVP.

• Specialised small models vs LLMs
• Data you already have
• Acceptance criteria
• On-premise and edge deployment

Technology: Physical AI →

2 days, hands-on - ML, software and quality engineers

Video & Image Recognition

Detect, segment and classify what matters in your screens, scans and camera streams with a fine-tuned vision-language model.

• Detect, segment, classify
• Labelling your own images
• Fine-tuned vs base model
• Evaluation per screen or site

Technology: Video & Image Recognition →

1 day, hands-on - ML and embedded engineers

Edge AI: Models on the Device

Make a model fit the compute already in the camera, machine or vehicle, and keep it working without a network.

• Quantisation and optimisation
• Fitting the hardware you have
• Real time without a connection
• Regression on real devices

Technology: Edge AI →

1 day - Automation and process engineers

Digital Twins & Simulation

Use a virtual copy of your line, warehouse or robot cell to train and test models before anything changes on the real floor.

• Modelling layout and material flow
• Synthetic scenes and rare cases
• Trying changes in the twin first
• Hardware-in-the-loop validation

Technology: Digital Twins & Simulation →

Half day, on site - Operators, line leads, maintenance

Cobot Operation & Safety

Working next to a vision-guided cobot: safe operation, exception handling and daily checks.

• Safety zones and stops
• Handling exceptions the robot flags
• Daily checks and hand-over
• When to call support

1 day, hands-on - Data and ML engineers

Dataset Processing

Turn your own screens, images, documents and logs into training-ready datasets that stay on your site.

• Collecting from your systems
• De-identification
• Labelling and ambiguous samples
• Dataset versions and traceability

Technology: Dataset Processing →

2 days, hands-on - ML and software engineers

Fine-Tuning & Deploying Small Models

Fine-tune a vision-language model on your own screens or images, evaluate it and deploy it to edge hardware.

• Dataset processing and de-identification
• Fine-tuning workflow
• Reproducible evaluation
• Quantisation and rollout

Technology: Model Fine-Tune & Deploy →

1 day - Planners, maintenance leads, data analysts

Predictive AI: Forecasts from Your History

Forecast failures, demand and delays from your sensor logs, service history and orders, early enough to act.

• Failure prediction
• Demand and stock forecasts
• Confidence levels
• Measuring against real outcomes

Technology: Predictive AI →