匠人学院 JR Academy
OpenAI Select PartnerClaude Certified Services Partner

FDE Tech · Engineering delivery

FDE Tech Career Advancement

Course duration
10 weeks
First-cohort preorder
A$4,400

FDE training for career targets of A$200,000+ in Australia and US$250,000+ in North America

Build on your AI Engineering skills to take on FDE client solutions and delivery. We give you the company context, existing systems and budget constraints. You investigate, propose a solution and present it to mentors and peers, then revise and validate it in response to questions before making an evidence-backed client pitch.

For engineers who can already build AI applications and want to move into FDE client delivery.

Proposal & review
These figures describe target roles, not a guarantee of employment or salary.

Working mainly in business, product or project delivery? Explore FDE Business

Why train for FDE now?

In 55 seconds, see what forward deployed engineers do and how you’ll practise over 10 weeks. English subtitles are available.

We bring the company brief. You bring the proposal and pitch.

You do not need prior business ownership. Watch how discovery and trade-offs work, then progressively own interviews, proposals, pitches and delivery.

Each round starts with a company brief. Form your own judgement, present a proposal, and defend it in front of the group. Use the questions and new constraints to revise both your solution and your delivery plan.

Client Brief · Understand the company

Your work
Read the company context, systems, workflows, budget and constraints. Identify missing information and questions for the client.
Your output
Discovery questions, current workflow and project goals.

The review question:Who has the problem, who decides, and which assumptions need evidence?

What this round must establish

Which business problem is worth solving?

Investigate the context, goals, workflow, data and constraints. Separate facts from assumptions and ask what is missing.

What you bring to the review

Discovery brief, problem definition and system boundaries

Who responds or challenges · Mentor provides context / peers act as business owners

Respond to questions and revise the proposal, then repeat with another company and different constraints. Finish with a complete client pitch covering the problem, solution, evidence and investment. Each participant explains their own decisions and contribution.

Ten dimensions of FDE delivery

Defend decisions with evidence and adapt when conditions change. Review uses individual artefacts, observed work and contributions.

Business Acumen

Understand revenue, costs and whether a problem merits investment.

Client Discovery

Investigate unfamiliar operations, evidence, baselines and decision authority.

Solution Architecture

Compare architectures and explain choices under client constraints.

Commercial Planning

Plan scope, full costs, milestones, dependencies and stopping conditions.

AI Governance

Implement data residency, access, security, approval and accountability controls.

Engineering Execution

Integrate with existing systems and handle retries, duplicates and recovery.

AI Reliability

Build golden data and evaluate quality, cost, latency and failure modes.

Stakeholder Leadership

Coordinate business, IT, security and budget decisions through conflict and change.

Executive Communication

Explain investment, risks and evidence; answer objections and seek approval.

Enterprise Adoption

Enable users, assess pilots and verify independent handover and operation.

  1. 01 · Facilitator-led
  2. 02 · With a guide
  3. 03 · Independent
  4. 04 · Adaptive

Build an enterprise AI project you can demonstrate and evaluate

Bring the proposal and implementation together: demonstrate the key workflow, then explain your choices with architecture, budget and evaluation records. Use this evidence in interviews and project discussions.

Customer proposal, architecture and budget

Problem definition, solution comparison, architecture and cost, with explicit scope, dependencies and acceptance criteria.

Code, integrations and evaluation

A working core workflow with system integration, permissions and recovery, supported by quality, latency and cost evidence.

Customer presentation and operating handover

Presentation, decision and change records, plus usage guidance, a runbook and maintenance ownership.

PROPOSAL → REVIEW → REVISION

A proposal must stand up to scrutiny

Start with a client brief, propose a solution, present it, answer questions and revise the plan. Read both complete versions of the Knowledge Access course case below.

Knowledge Access · Enterprise Policy Assistant

40 staff across two departments need faster access to current policy, within their existing permissions, budget and approval boundaries.

Revised decision: a bounded internal pilot, with A$21,600 build and A$1,060 estimated monthly running costs.

  • A$24,000 build ceiling
  • A$1,200 monthly running ceiling
  • Four-week client pilot
  1. 01Discover the problem
  2. 02Propose a solution
  3. 03Present and review
  4. 04Revise and pitch

Scope follows the client decision

First draftThe unapproved archive remains part of the draft coverage.

Review question
How can an unapproved interface be included in the delivery scope?

Revised decisionRemove the archive connector, remeasure coverage and reassess cost and acceptance if approval arrives.

Permissions cannot depend on a prompt

First draftUsers supply their own department and the model is asked to enforce confidentiality.

Review question
What happens if an operations employee enters finance as their department?

Revised decisionDerive identity from the server session, filter before retrieval and check revocation and caching.

Budget for the complete delivery

First draftA$14,880 includes discovery, development and contingency only.

Review question
Who performs evaluation, pilot preparation and handover, and how long do they need?

Revised decisionAdd 32 evaluation hours and 24 handover hours: A$21,600, within the ceiling.

Value must survive a finance review

First draftA$19,200 capacity value is labelled cash savings; maintenance is missing.

Review question
If payroll has not fallen, has this cash actually been saved?

Revised decisionSeparate capacity from cash, test coverage sensitivity and include A$1,060 monthly running costs.

Average accuracy can hide a critical failure

First draftThe draft asks only for 80% answer accuracy.

Review question
Can you launch if generally useful answers leak one finance document?

Revised decisionBlock on access leaks and unauthorised execution; check citations, refusals and escalation separately.

A repository is not a handover

First draftDeliver code and a README, assuming IT can maintain it.

Review question
Who responds to a source update or access incident, and how do they recover?

Revised decisionSpecify owners, versions, rollback and incident response, then have the receiver run the checks.

What does the mentor review examine?

Present your own reasoning. Mentors question the evidence and trade-offs; peers take business, IT, security and budget perspectives. Explain how new constraints change your decision.

01 · Proposal Review

Present the Discovery Brief, architecture, scope and budget. Defend the opportunity, chosen approach and unresolved assumptions.

02 · Technical Review

Open the implementation and support your reasoning with access checks, failure records and Evals. Match the proposal’s commitments to observable behaviour.

03 · Client Pitch after Change

Revise the proposal and implementation after review. Explain changes to scope, cost and risk, the decision requested and subsequent ownership.

Mentor team

Lightman

Lightman

JR Academy founder

Startup experience, full-stack development and product design.

Joe 韭神

Joe 韭神

Former Googler SDE · Block SDE

Former Google software engineer (SDE), with SDE experience at Block; also mentors in software engineering and careers.

Jason

Jason

FDE experience · V2

Forward Deployed Engineer experience at V2 and an AI Engineer VIP mentor, supporting AI engineering projects and technical mentoring.

Roy

Roy

FDE experience · V2

Forward Deployed Engineer experience at V2, with a focus on applied AI and AI engineering and a software engineering background.

Share your project experience and target role to discuss the mentor assignments and training arrangements for the cohort.

What are their classes like? | Roy, Jason and Lightman classroom highlights

Watch on YouTube

What changes when you take on client delivery?

Building a feature also brings business questions: Is this worth doing? Who can approve it? What do existing systems allow? Who will use and maintain it?

AI Engineering is your starting point

  • Build and debug AI applications and tool calls.
  • Work with data, evaluate outputs and diagnose failures.

FDE adds client delivery responsibility

  • Turn an unclear business request into an engineering plan.
  • Make trade-offs across budgets, access and stakeholders.
  • Build with the client, hand over the system and explain results.
BIG TECH. ENTERPRISE AI. SENIOR OWNERSHIP.

Benchmark the FDE roles at major tech companies

Build towards enterprise AI delivery roles in Australia and North America.

Australia · A$200,000+ annual salary target

For engineers aiming to own enterprise AI solutions: discover the problem, propose the architecture, plan the budget, and take responsibility for implementation and acceptance.

See how we train these capabilities

United States · Published compensation references

For experienced engineers aiming at enterprise AI delivery. Explore US Google, Atlassian, Figma and Salesforce roles by level, with base pay and reported total compensation shown separately.

US references: experienced engineering roles, with level and location shown. Employer-posted ranges and reported level averages are individually attributed.

USD · Your comparable annual pay

View source: Google Cloud
Google Cloud · Forward Deployed Engineer IV · Applied AI · US

Annual difference from this benchmark+$7,000
+$100,000

Difference as a percentage of your input
+3.5% → +50.0%

Compare the same pay basis. Total pay includes reported equity and bonus; share values and vesting affect its realised value. Differences are comparisons, not an offer or course outcome.
How to read these figures

Posted ranges are role-specific: Google IV lists a 20% bonus target plus equity outside base; Atlassian uses Zone A; Figma specifies Senior/Staff. Salesforce Lead MTS base is a reported level average. Total-pay medians use the FDE-specific Levels.fyi filter: Google17 submissions, Palantir229; Salesforce does not disclose the summary sample count. These mixed-level company samples are not pooled into one big-tech average.

What changes when you take on enterprise AI delivery?

Compare the responsibilities below with your experience. Select one to see the course practice that develops it.

Week 01–03

Define the work before the build

Clarify the customer problem, define scope and sequence delivery.

In this course

Question the enterprise brief, surface missing information and defend a bounded proposal.

Proposal → Pre → Pitch
Additional reference: frontier AI labs

These individual employer ranges are additional references, not the market average. Staff level and pay definitions remain role-specific.

OpenAI

Forward Deployed Software Engineer · SF

$185,000–$325,000

Base pay · equity additional

View source ↗

Anthropic

Staff Forward Deployed Engineer · NY / SF / Seattle

$280,000–$320,000

Annual Salary · official label

View source ↗

Role reference: OpenAI · Forward Deployed Engineer · SydneyOpenAI · Forward Deployed Software Engineer · SF · Anthropic Staff FDE

Checked: 2026-11-04

How the 10-week Tech course progresses

Observe and participate in weeks 1–2, practise with a guide in weeks 3–4, then own the main project in weeks 5–10. Budget, governance, evaluation and client decisions recur throughout.

10 weeksCourse duration
3 hoursWeekly workshop
3–5 hoursWeekly independent practice
Week 1

Enterprise Discovery & AI Opportunities

Observe a facilitator-led discovery, map how a company completes a service request, then ask follow-up questions.

Stage deliverableWorkflow, stakeholder map, evidence and unanswered questions.
Week 2

Solution Design & Delivery Planning

Compare process changes, code automation, RAG and agents; pitch scope, full costs, dependencies and stopping criteria.

Stage deliverableProposal, alternatives, architecture, budget and milestones.
Week 3

AI Governance, Data & Security

Interview business, IT and security roles; turn data residency, access and approval requirements into design and implementation controls.

Stage deliverableData flows, access controls, approvals and technical validation.
Week 4

AI Evaluation & Reliability

Build business-reviewed golden data, separate development and acceptance sets, analyse failures and defend the pilot decision.

Stage deliverableGolden dataset, scoring criteria, failure analysis and revised proposal.
Week 5

Client Discovery & Solution Proposal

Own discovery for the main project and seek approval for the proposal, budget, acceptance criteria and milestones.

Stage deliverableProject proposal, decision log, budget and acceptance plan.
Week 6

Enterprise Integration & Implementation

Implement against existing systems; handle least privilege, environment boundaries, duplicate events and long-running work.

Stage deliverableWorking core flow, interface contracts and integration evidence.
Week 7

Evaluation, Cost & Performance

Evaluate bounded execution, human takeover and recovery; measure quality, latency and operating costs before changing the implementation.

Stage deliverableEvaluation report, cost and performance evidence, recovery checks.
Week 8

Stakeholder Management & Change Negotiation

Negotiate budget cuts, interface delays and added requirements; revise scope, architecture, cost and schedule.

Stage deliverableChange request, impact analysis, revised proposal and decision log.
Week 9

Pilot Delivery & User Enablement

Use baselines and pilot evidence to decide investment; teach users to operate, review and recover the system.

Stage deliverablePilot decision, user instructions, enablement and rollback plan.
Week 10

Acceptance, Handover & Final Pitch

Defend acceptance evidence and individual decisions; let another participant operate and recover the project using the handover pack.

Stage deliverableFinal system, acceptance matrix, runbook, handover pack and pitch.

Confirm cohort dates and session times before enrolment. Any foundation self-study is additional.

30 engineering tasks to take AI applications into the enterprise

Once an AI app works, enterprise delivery adds legacy systems, access restrictions, budgets and changing requirements. These 30 tasks span six delivery areas and combine across multiple cases over 10 weeks, producing proposal, implementation and validation evidence.

Discovery · 5Architecture · 5Integration · 5Agents · 5Evaluation · 5Delivery · 5

Discovery & Delivery Planning5

Understand the customer environment before committing engineering effort.

01 / 30

Enterprise System Discovery

Inspect an unfamiliar repository, interface documentation and operating records to trace a business request and identify gaps.

DeliverableSystem boundary map, interface inventory and investigation questions.

02 / 30

Discovery under Uncertainty

Investigate incomplete customer requirements and distinguish evidence, assumptions and unknowns about users, decision-makers and constraints.

DeliverableDiscovery brief and explicit assumptions.

03 / 30

Data Readiness and Ownership

Check schemas, missing values, versions, ownership and access conditions to decide what data can enter the first phase.

DeliverableData readiness assessment and ownership boundaries.

04 / 30

Baselines and Acceptance

Turn requests for better accuracy or speed into inspectable requirements, a baseline and an initial set of checks.

DeliverableBaseline, acceptance criteria and initial check set.

05 / 30

Scope, Budget and Delivery

Estimate discovery, implementation, integration, model usage, review and maintenance, including dependencies and stopping conditions.

DeliverableScope, budget assumptions and phased delivery plan.

Architecture & Trade-offs5

Make architecture choices against data, access, quality and cost constraints.

06 / 30

Choosing the Right Architecture

Compare deterministic automation, RAG and agents for the same requirement, using constraints and validation evidence.

DeliverableArchitecture comparison and decision record.

07 / 30

Model Routing and Cost Control

Compare models by task quality, response time and cost, then validate routing, caching and spending limits.

DeliverableModel comparison, routing strategy and cost estimate.

08 / 30

Retrieval with Evidence

Build retrieval with chunking, versions and citations, including explicit handling of insufficient evidence.

DeliverableRetrieval implementation, citation checks and insufficient-evidence handling.

09 / 30

Identity-aware Retrieval

Enforce user, department and tenant access in retrieval and check whether authorisation actually constrains results.

DeliverableAuthorised retrieval and boundary-access checks.

10 / 30

Context and Knowledge Lifecycle

Handle long context, updates, stale sources and removal requirements with clear retrieval and invalidation decisions.

DeliverableContext strategy, update flow and knowledge lifecycle rules.

Enterprise Integration5

Connect existing interfaces and data under operational constraints.

11 / 30

Working with Legacy APIs

Adapt existing APIs with inconsistent fields, versions, rate limits and error responses.

DeliverableAPI adapter, contract checks and error mapping.

12 / 30

MCP in Enterprise Systems

Define tool inputs, outputs, identity and access boundaries, checking exposure against the actual business need.

DeliverableMCP adapter, parameter validation and access contract.

13 / 30

Events, Idempotency and Duplicates

Handle repeated messages, callbacks and out-of-order events without duplicating business writes or actions.

DeliverableEvent handler, deduplication record and repeat-input checks.

14 / 30

Long-running Workflows

Track long-running work through timeouts, failed-job queues, bounded retries and recovery.

DeliverableJob state, recovery paths and failure-handling code.

15 / 30

Deployment and Environment Boundaries

Separate configuration, secrets, data and runtime environments so the same code starts under explicit, verifiable conditions.

DeliverableDeployment configuration, environment checks and startup instructions.

Agents & Workflow Control5

Bound tool use and make workflows stoppable, inspectable and recoverable.

16 / 30

Bounded Agent Execution

Define steps, tools, time and cost limits, and verify when an agent continues, pauses or stops.

DeliverableAgent state machine and execution-boundary checks.

17 / 30

Approval-aware Tool Use

Require approval for consequential actions and check whether it remains valid when conditions change.

DeliverableApproval state, execution checks and changed-condition records.

18 / 30

Skills for Team Workflows

Package validated operations as reusable skills with input requirements, tool access, versioning and failure handling.

DeliverableReusable skill, usage contract and version checks.

19 / 30

Human Handover and Resume

Preserve state for human decisions, then resume from the correct step without repeating completed actions.

DeliverableHandover record, checkpoints and resume implementation.

20 / 30

Recovery after Partial Failure

Reconcile partially completed actions before choosing compensation, retry or human intervention.

DeliverableReconciliation flow, compensation strategy and recovery checks.

Evaluation & Reliability5

Use quality, access, latency and cost evidence to judge delivery readiness.

21 / 30

Evals and Regression Gates

Build evaluations from normal inputs, boundaries and known failures, with regression criteria for model, prompt and tool changes.

DeliverableEvaluation set, comparison report and regression criteria.

22 / 30

Prompt Injection and Trust Boundaries

Check whether instructions from documents, email and tool results cross trust boundaries or are mistaken for authority.

DeliverableAdversarial inputs, observed outcomes and trust-boundary analysis.

23 / 30

Testing Retrieval and Permissions

Evaluate citations, knowledge versions, abstention and cross-user access as distinct retrieval and authorisation concerns.

DeliverableRetrieval and permission evaluation with failure analysis.

24 / 30

Latency and Unit Economics

Compare latency, model spend and human review across workloads, estimating cost per accepted outcome.

DeliverableLatency measurements, unit-cost analysis and optimisation trade-offs.

25 / 30

Observability and Incident Diagnosis

Trace tools, errors, costs and decisions across a request, retaining useful evidence while controlling sensitive log content.

DeliverableTraces, alert conditions and incident diagnosis records.

Customer Delivery & Handover5

Handle change with working evidence, then prepare the pilot and handover.

26 / 30

Technical Reviews with Stakeholders

Review architecture, implementation and evaluations with business, IT, security and budget stakeholders.

DeliverableCustomer presentation, decision record and open issues.

27 / 30

Engineering through Change

Respond to budget cuts, delayed interfaces or new requirements by revising scope, architecture and implementation.

DeliverableChange record, revised code and updated budget and plan.

28 / 30

Release Readiness and Rollback

Check release conditions, versions, monitoring and ownership, and plan containment, rollback and recovery.

DeliverableRelease checklist, version record and rollback plan.

29 / 30

From Pilot to Adoption

Bound the pilot and use baseline comparisons, operational issues and review cost to decide whether to continue, adapt or stop.

DeliverablePilot analysis and adoption plan.

30 / 30

Handover that Others Can Run

Prepare operating, maintenance and incident guidance, then check whether another participant can run and diagnose the key workflow.

DeliverableRunbook, maintenance ownership and handover package.

Put your ideas in front of other people

FDE practice brings presentations, questions and hands-on validation together. These photos are from previous JR Academy classes and events.

An instructor and learners examining work on a laptop in a previous JR Academy class
Previous classes · Working through problems together
A speaker and participants talking at a previous JR Academy event
Previous events · Sharing ideas and taking questions

Six modules. One complete delivery discipline.

Observe and participate in weeks 1–2, practise with a guide in weeks 3–4, then own the main project in weeks 5–10. Budget, governance, evaluation and client decisions recur throughout.

01 / 06

Understand the enterprise and qualify the problem

Enterprise Discovery

Observe a facilitator-led discovery, map how a company completes a service request, then ask follow-up questions.

Your deliverable

Workflow, stakeholder map, evidence and unanswered questions.

02 / 06

Connect architecture, scope, budget and delivery

Solution Architecture & Commercial Planning

Compare process changes, code automation, RAG and agents; pitch scope, full costs, dependencies and stopping criteria.

Your deliverable

Proposal, alternatives, architecture, budget and milestones.

03 / 06

Turn data, safety and reliability into delivery controls

AI Governance & Reliability

Interview business, IT and security roles; turn data residency, access and approval requirements into design and implementation controls.;Build business-reviewed golden data, separate development and acceptance sets, analyse failures and defend the pilot decision.

Your deliverable

Data flows, access controls, approvals and technical validation.;Golden dataset, scoring criteria, failure analysis and revised proposal.

04 / 06

Integrate, implement and evaluate in existing systems

Enterprise Engineering

Own discovery for the main project and seek approval for the proposal, budget, acceptance criteria and milestones.;Implement against existing systems; handle least privilege, environment boundaries, duplicate events and long-running work.;Evaluate bounded execution, human takeover and recovery; measure quality, latency and operating costs before changing the implementation.

Your deliverable

Project proposal, decision log, budget and acceptance plan.;Working core flow, interface contracts and integration evidence.;Evaluation report, cost and performance evidence, recovery checks.

05 / 06

Negotiate conflict and change to reach a decision

Stakeholder Leadership

Negotiate budget cuts, interface delays and added requirements; revise scope, architecture, cost and schedule.

Your deliverable

Change request, impact analysis, revised proposal and decision log.

06 / 06

Enable users and verify independent handover

Enterprise Adoption

Use baselines and pilot evidence to decide investment; teach users to operate, review and recover the system.;Defend acceptance evidence and individual decisions; let another participant operate and recover the project using the handover pack.

Your deliverable

Pilot decision, user instructions, enablement and rollback plan.;Final system, acceptance matrix, runbook, handover pack and pitch.

Study enterprise AI through ANZ, Woolworths and IAG

Start with products these companies describe publicly, then examine their delivery challenges. Classroom extensions introduce budget, data and requirement changes for you to work through.

Enterprise case analysis · ANZ · BANKING

ANZ · Z-GPT: connecting internal bank knowledge

What the company has published

ANZ described Z-GPT in 2023: an internal chatbot using OpenAI models through Microsoft, with experiments on subsets of internal data. Users were responsible for reviewing answers and making the final decision.

Read the official source · ANZ Bluenotes · Z-GPT ↗
Public product
Z-GPT
Users
Internal staff
Source date
2023-08-22
The delivery question

For a comparable bank assistant, how would you define source access, identity boundaries and answer acceptance criteria?

Classroom analysis: choosing an approach

Start the classroom design with bounded internal Q&A, source checks and human responsibility before considering expansion.

Classroom solution walkthrough

  1. Authenticate
  2. Permission-aware retrieval
  3. Check version & citation
  4. Answer or escalate
Classroom analysis: budget and scope

Compare existing enterprise tools, retrieval and custom integrations. Include data governance, integration, evaluation and adoption effort.

A change introduced in training

Training condition: security rejects a source. Revise the pilot scope and budget recommendation for the sponsor.

Tech: what you build

Identity and retrieval access checks, version filters, citation and abstention evals, per-request costs and exception logs.

One request. Different constraints. Different solutions.

Start with the work and the client’s constraints. Across the cases, data, access and budgets change—and you need to explain how your choice changes with them.

What does this work actually need AI to do?
Direction · Automate the predictable parts with rules or code

What to validate · Exception ownership and maintainability when rules change.

Official Claude resources, used when you need them

Revisit APIs, Claude Code, MCP or Skills where you need preparation. Use delivery resources when planning a client pilot. Official page previews and summaries are available here.

Claude’s official FDE learning path
Claude’s official FDE learning path · Source: Anthropic Partner Academy
Official course: Delivery Methodology
Official course: Delivery Methodology · Source: Anthropic Partner Academy
Before class: fill specific gaps

Use these courses where you need them. You still need to build and debug AI applications independently before joining.

  • AI Capabilities and Limitations

    Recognise model limitations and explain them to users.

  • AI Fluency for Builders

    Practise delegation, review and developer responsibility.

  • Claude Platform & Model Foundations

    Consider entry points, features, models and context.

  • Choosing the right Claude model

    Compare models for a task and its usage limits.

  • Claude Platform 101

    Review APIs, tool calls and agent loops.

During projects: look up engineering techniques

Return to these materials when working in a client repository or connecting business systems. Apply what you find in your project.

  • Claude Code 101

    Explore, plan, implement and check changes in a repository.

  • Claude Code in Action

    Manage permissions and verification for long tasks.

  • Introduction to Model Context Protocol

    Connect tools and data through MCP.

  • Introduction to agent skills

    Package repeatable operations into reusable Skills.

During delivery: plan a pilot people can use

Use the official pilot and adoption guidance when preparing scope, budget, acceptance criteria and handover for the class case.

  • Delivery Methodology

    Qualify a use case, design a pilot and capture a baseline.

  • Driving enterprise adoption of Claude

    Plan user activation, adoption and expansion.

  • Enterprise Integration & Production

    Further reading on production integration.

  • Responsible AI, Safety & Risk for Architects

    Further reading on architecture safety and risk.

FDE Tech · Standalone enrolment

First-cohort presale · 30 places

22 places remaining8 enrolled

Regular course price

A$4,800
First-cohort preorder
A$4,400
Early bird
A$4,600

Tech is a standalone engineering delivery course. Regular price A$4,800; first-cohort preorder A$4,400; early bird A$4,600. No prior JR Academy AI Engineer purchase is required. Entry depends on engineering skills and project experience.

  • 30 enterprise engineering tasks across multiple customer cases
  • Proposal, budgeting and presentation practice
  • Six delivery modules and official Claude learning resources
Confirm preorder and early-bird eligibility, cohort dates and support arrangements when enquiring.

Is this the right next step for you?

Consider the course if you already build and debug AI applications independently and want responsibility for enterprise requirements, solutions and delivery.

  • Run and explain your own AI project.
  • Modify code, diagnose failures and verify results.
  • Set aside time for project work and presenting to classmates.

A separate course · Separate enrolment

FDE Business Career Advancement

Moving towards FDE work from business, product or consulting? Build AI technical fluency, automation practice, discovery, proposals, pitches and client-delivery evidence.

Regular A$3,850 · First-cohort preorder A$3,500 · Early bird A$3,650

Explore FDE Business

Questions before you enrol

Can I join without taking JR Academy’s AI Engineer course?

Yes. Tech requires the ability to build, debug and explain your own AI applications; purchasing JR Academy AI Engineer first is not required. The two FDE courses also have separate enrolment.

I know the business but cannot build AI applications independently. Which course fits?

Explore Business first: it builds on domain experience with AI automation, proposals, budgets and cross-team delivery. Tech requires independent AI application development and debugging. Specific gaps in MCP, Skills or other tools can be addressed with official resources.

Explore FDE Business

How much time should I set aside?

The current design is 10 weeks, with a 3-hour weekly workshop and around 3–5 hours of project and presentation preparation. Confirm actual cohort dates and session times before enrolment.

How does this differ from Business training and one-to-one coaching?

Tech requires independent AI implementation. Business is a separate course focused on business judgment, proposals, budgets and adoption. They have separate enrolment. One-to-one coaching addresses an individual background and project.

What role do the official Claude courses play?

Use official resources to fill specific gaps in models, APIs, Claude Code, MCP and Skills before class, and consult their pilot and adoption methods during cases. JR Academy organises the main workshops around solution design, implementation, evaluation and customer delivery.

Are the salary targets guaranteed outcomes?

No. A$200,000+ in Australia and US$250,000+ in North America describe career targets. Outcomes depend on experience, skills, interviews and market conditions. The course does not guarantee employment or salary.

Make your next step specific

Tell us what you do, which AI projects you have built and the roles you want. Discuss fit before deciding whether to enrol.