Contact us

AI-Native SDLC Training

Learn to take a product from idea to production with a coding agent while keeping control of every decision, following Anthropic's AI-Native SDLC playbook. You leave with a method to extend these practices across your organisation.

Training objectives

  • Frame a need with a coding agent through to a validated specification
  • Equip an agent with the guidelines, methods and deterministic controls that govern and verify its work
  • Decide on a merge from a review, then authorise a production release
  • Diagnose a production incident with a coding agent
  • Make your coding agent's work reliable and measure its effect at each phase of the cycle
  • Design the common foundation of tools, skills and hooks to share across your organisation

Who should attend?

This course is for Product and Tech teams (PM, PO, Product Designers, architects, developers, QA, Ops, platform...) and the people leading their transformation, who want to rethink their development lifecycle in the age of AI.

Use an LLM daily in your work.

Programme de formation

Training program

Download the program

Heading 1

with a request body that specifies how to map the columns of your import file to the associated CRM properties in HubSpot.... In the request JSON, define the import file details, including mapping the spreadsheet's columns to HubSpot data. Your JSON request should include the following fields:... entry for each column.

Module 1 - Introduction to the AI-Native SDLC playbook

  • Identifying the phases that become the bottleneck when an agent writes the code
  • Distinguishing a linear, role-based cycle from a loop where each accepted artefact starts the next phase
  • Locating what governs an agent's work: context, skills, hooks, sub-agents, evals and approvals
  • Mapping the twelve practices of the playbook, their phases and their dependencies
  •  

Module 2 - Plan : framing the need

  • Connecting an agent to a code repository and giving it a reusable framing method
  • Clarifying a need with the agent while keeping control of the product choices
  • Checking that a framed need is ready for design, with no decision invented by the agent
  • Tracking the time taken to capture a need and the share of needs accepted into design

Module 3 - Design : specifying the solution

  • Having a specification written as verifiable behaviours, following the organisation's standards
  • Reviewing the requirements and deciding, as Product Owner, the questions the agent leaves open
  • Approving the specification through a pull request
  • Tracking design lead time and rework of the specification once development has started

Module 4 - Build : planning and writing the code

  • Having the agent prepare the project guidelines, separating what guides it from what enforces a rule
  • Assessing an implementation plan by its risks and the evidence it provides before authorising the code
  • Choosing between a parallel session and a sub-agent depending on the task
  • Tracking the gap between the accepted plan and the delivered code, and the mistakes the agent repeats

Module 5 - Test : getting the work verified

  • Grouping the project's checks into single commands whose result counts as proof
  • Having the test that reproduces a bug written, protecting it during the fix, then reviewing the diff
  • Distinguishing tests, which verify the product, from evals, which verify the agent
  • Tracking first-pass CI success, eval success and the time taken to turn an incident into an eval

Module 6 - Deploy : governing review and delivery

  • Giving the review to a session that did not write the code and getting findings ranked by severity
  • Grading the agent's autonomy by environment, up to the authorisation of a production release
  • Preparing a delivery file: conditions met, conditions still to be tested, rollback plan
  • Tracking waiting time at approvals, defects caught before merge and violations in production

Module 7 - Maintain : closing the loop

  • Detecting a threshold breach with a script that uses no AI and sets each level's action in advance
  • Getting a diagnosis from the agent that separates facts, hypotheses and open questions
  • Recording the breach and its diagnosis in a new need that opens the next cycle
  • Tracking the time between a breach and the new need, and the recurrence of incidents of the same class

Module 8 - Beyond the playbook : building a common foundation

  • Comparing the artefacts and sequence of BMad and OpenSpec with those of the cycle practised
  • Identifying the strengths of five market tools, including Spec Kit, gstack and Superpowers
  • Deciding for each building block whether to adopt, adapt, encapsulate or build it
  • Defining the Golden Path, the organisation's common foundation: one skill per phase and control hooks
  • Splitting the agent's tooling between what the repository carries and what the platform provides

Interested?

Pre-registration

Pre-registration

For your teams

from 10,400€ exl. VAT

Private session, for your own company (up to 10 people)

Level : Intermediate Duration : 2 days
Next sessions : 
Contact the Academy teams
Schedule a private session

Our training courses on the same topic

Compare this course with another one in its category! Who knows, maybe you’re missing out on a hidden gem?

Heading 1

with a request body that specifies how to map the columns of your import file to the associated CRM properties in HubSpot.... In the request JSON, define the import file details, including mapping the spreadsheet's columns to HubSpot data. Your request JSON should include the following fields:... entry for each column.

Data AI

AI-Augmented PM Training (Claude)
Duration : 2 days
Level : Intermediate

Take the AI turn across the entire Product cycle: map your uses, get hands-on with Claude's features, then Claude Code to create your own skills, connect your tools and build a first application on your own use case. 

Read more
Data AI

AI-Augmented PM Training (GitHub)
Duration : 2 days
Level : Intermediate

Take the AI turn across the entire Product cycle: map your uses, get hands-on with GitHub Copilot features to create your own skills, connect your tools and build a first application on your own use case. 

Read more
Data AI

AI-Augmented PM Training (ChatGPT)
Duration : 2 days
Level : Intermediate

Take the AI turn across the entire Product cycle: map your uses, get hands-on with ChatGPTs features, then Codex to create your own skills, connect your tools and build a first application on your own use case. 

Read more
trainers

Our Thiga trainers

Our training courses are led by one of our experts. They are simultaneously trainers, program creators and consultants and they combine their professional experience and pedagogy to help you develop your skills.

trainers
Discover our trainers

Heading 1

with a request body that specifies how to map the columns of your import file to the associated CRM properties in HubSpot.... In the request JSON, define the import file details, including mapping the spreadsheet's columns to HubSpot data. Your JSON request should include the following fields:... entry for each column.

A few words from David, Director of Thiga Academy

We are the leading Academy for Product training in France.
Our ambition? Drive the market maturity, help Product professionals grow, and assist companies in retaining their teams by offering career development paths, all while creating responsible digital products with impact.

A word from David, Director
of Thiga Academy

We are the benchmark Academy for product training in France.

91%

satisfaction

+11 000

alumni

+35

training

+1 200

certified

Contact us

Heading 1

with a request body that specifies how to map the columns of your import file to the associated CRM properties in HubSpot.... In the request JSON, define the import file details, including mapping the spreadsheet's columns to HubSpot data. Your JSON request should include the following fields:... entry for each column.

Read more
David-Schuster

David Schuster

Director of Thiga Academy since 2019

Frequently asked questions

All the answers to your questions about our training courses!

Are Thiga Academy training courses eligible for CPF?

Our training catalog is not eligible for CPF funding.

For freelancers or individuals wishing to train with us, please do not hesitate to contact us (academy@thiga.fr), and we will be able to offer you payment facilities.

How can I get my training financed?
To finance your training, there are two main criteria to take into account:
  • The professional situation of the people to be trained: employee, self-employed, job-seeker, private individual, etc.
  • The budget sources you can/want to mobilize: direct company funding, OPCO, France Travail, personal funds, etc.
Please note that our training catalog is not eligible for CPF funding, but we can offer you payment facilities. Please do not hesitate to contact us to discuss this.
Thiga Academy is Qualiopi certified in the "Training Action" category.
Why take this course?

Our industry-recognized training catalog offers a comprehensive program for every stage of your Product career.

This course is aimed at Product Managers, Product Designers, Product Marketing Managers, from Product Acculturation to the various levels of Product Leadership. We also offer to enhance your expertise with specialized training in AI Data, Accountability, Soft skills, ...

We run workshops on specific tools and methods used in day-to-day team work, and are also able to tailor programs to your needs.

More specifically, this program will enable you to learn to take a product from idea to production with a coding agent while keeping control of every decision, following Anthropic's AI-Native SDLC playbook. You leave with a method to extend these practices across your organisation.

For more information, you can contact our teams.

Does this training lead to certification?

Our AI-Native SDLC training course is not yet certifiable. But our entire catalog will soon be!

I want to customize this training

This training program meets the needs of your teams, but you'd like to adapt, replace or add a module... It's possible!

Click on "Schedule a private session" at the top of this page and tell us what you'd like to adapt.

More questions?

Would you like to train yourself or your team at Thiga Academy? Contact our teams today!

Contact us

Heading 1

with a request body that specifies how to map the columns of your import file to the associated CRM properties in HubSpot.... In the request JSON, define the import file details, including mapping the spreadsheet's columns to HubSpot data. Your JSON request should include the following fields:... entry for each column.