Centaur CareersCentaur Careers

CURRENT PROGRAMME / APPLIED AI

Applied AI
Engineering.

Machine learning / Deep learning / Agentic AI

Follow the signal from raw data to a working AI system. Explore the learning path, practise with project briefs and choose the role direction you want to build towards.

Explore six subjects Choose a role
01 Learn the foundations02 Build and test systems03 Explore a role direction

OPEN A SUBJECT / 01 TO 06

Six subjects.
One connected build.

Choose any subject to see its own visual explanation, practical sequence and example work. Each chapter belongs to Applied AI Engineering.

01 / INPUTData foundationsMake data ready for a question.Open subject 02 / TRAINMachine learningTrain a model you can explain.Open subject 03 / LEARNDeep learningWork with neural systems.Open subject 04 / GROUNDLLM applicationsConnect answers to their sources.Open subject 05 / ACTAgentic systemsCoordinate tools with oversight.Open subject 06 / SHIPProduction practicePackage the work for review.Open subject

THE IDEA / A BUILD YOU CAN EXPLAIN

Learn the system.
Show the work.

AI roles ask for different kinds of evidence. Follow a connected build sequence and see what a role-aligned project could look like.

Explore the learning path
THE EVIDENCE SEQUENCE01 / 04

INPUT

Start with a question.

Choose a problem, inspect the available data and make the starting assumptions explicit.

WHAT TO SHOWA clear brief and a reproducible starting point.

01 / THE LEARNING PATH

One signal.
Six transformations.

Move through the six learning stages to see what each part of the system asks you to build and explain.

01 / 06

  1. 01 / INPUT

    Data foundations

    Make data ready for a question.

    Start with Python, structured data, SQL, basic statistics and reproducible work.

    THE WORKInspect, clean and document a fictional dataset.
    THE EVIDENCEA reproducible data notebook with clear assumptions.
    Explore this subject
  2. 02 / TRAIN

    Machine learning

    Train a model you can explain.

    Move through features, train and validation splits, leakage, baseline models and meaningful evaluation.

    THE WORKCompare a simple baseline with a trained model.
    THE EVIDENCEAn evaluation report that includes failure cases.
    Explore this subject
  3. 03 / LEARN

    Deep learning

    Work with neural systems.

    Explore neural networks, transfer learning and focused text or image tasks.

    THE WORKTest a model on a fictional classification task.
    THE EVIDENCEA model card with results and limitations.
    Explore this subject
  4. 04 / GROUND

    LLM applications

    Connect answers to their sources.

    Design retrieval, source attribution, prompt workflows and evaluations for an LLM application.

    THE WORKBuild a document assistant using an invented corpus.
    THE EVIDENCEA source-grounded assistant and test set.
    Explore this subject
  5. 05 / ACT

    Agentic systems

    Coordinate tools with oversight.

    Explore tool calls, state, human approval, traces and failure handling in a bounded workflow.

    THE WORKSimulate a multi-step task that pauses for approval.
    THE EVIDENCEA workflow trace with safeguards and failure tests.
    Explore this subject
  6. 06 / SHIP

    Production practice

    Package the work for review.

    Bring together an API, deployment notes, monitoring questions and cost awareness.

    THE WORKPresent a working prototype and its limits.
    THE EVIDENCEA demonstrable service and concise technical handover.
    Explore this subject

02 / ROLE DIRECTIONS

Pick a problem to own.

One foundation can lead to different kinds of portfolio work. Select a direction to see relevant skills and a sample artifact. Employers set their own requirements for each role.

ACTIVE DIRECTION01 / 05

Investigate / build / explain

Data Scientist

Turn a question and a dataset into a tested finding.

ExplorationExperiment designEvaluation
SAMPLE PORTFOLIO ARTIFACTA reproducible analysis with model comparisons and documented limitations.

Role directions help you plan portfolio work within this course. They are not job offers or separate qualifications.

03 / PROJECT STUDIO

Ideas become inspectable work.

These fictional practice briefs show the type of evidence a learner can create. They are examples, not claims about past student projects.

WORKBENCH / 01

Document triage

Can a model sort fictional documents and identify where a human should review?

Input
Invented document examples
Build
A classifier, baseline and error review
Evidence
Metrics, model card and failure cases
NO BLACK BOX

Build it. Test it. Explain where it breaks.

ASK ABOUT THE PROGRAMME

Build your next step.

Tell us which AI direction interests you. The team can help you understand the course and discuss how to join.

Read the course answers

COURSE ENQUIRY / NO ENROLMENT COMMITMENT

Talk to us about AI.

Leave your details and the AI direction you want to explore. The team can contact you about the current programme. Sending this form is not an application or payment.

Submitting asks our team to contact you about Applied AI Engineering. Read our privacy policy.

QUESTIONS / ANSWERS

Before you join.

How do I ask about joining?

Use the course enquiry form on this page. The team can discuss the current programme and the steps to join.

Are the role directions separate courses?

They are ways to plan portfolio work inside Applied AI Engineering. They are not separate courses or qualifications.

Do I need a programming background?

Tell the team about your experience through the course enquiry form. They can help you understand where to start in the learning path.

Does the finance placement guarantee apply?

No. The published finance placement guarantee belongs to the Financial Operations Masterclass and does not apply to Applied AI Engineering.

View the Financial Operations Masterclass