Masterclass · October 2026
What AI in Salesforce programmes delivers today
Coding agents, Agentforce, MCP and new cost models are changing how Salesforce programmes are built and run. I put them in context for service, field service and B2B sales, including where Salesforce is not the right answer.
- 10chapters
- 32articles
- DE · ENGerman and English
Deadlines
Deadlines that matter now
These deadlines arrive whether your AI project is ready or not. Knowing them lets you plan interfaces and AI labelling in time.
- Local Salesforce DX MCP Server reaches end of life
- AI Act: labelling duty for existing generative AI systems
- End of integrations with username and password →
- AI Act: high-risk obligations under Annex III
- AI Act: high-risk obligations for products under Annex I
Chapters
Ten chapters, from the basics to manufacturing
The chapters build on each other but stand on their own. If you only need to prepare a decision, start with chapter 6.
- Chapter 01:
Terms and map
The four layers of AI in a Salesforce programme, what separates vibe coding from agentic engineering and which product name stands for what. - Chapter 02:
Salesforce tools
Which AI tools Salesforce itself offers for development and administration, what they cost and where their limits are. - Chapter 03:
MCP and interfaces
Which MCP servers and APIs connect AI agents to Salesforce, who reads which data through them and what the calls will cost. - Chapter 04:
External coding agents
Claude Code, Codex, Copilot and DevOps agents compared: what the tool decides and what contract, data residency and rules decide. - Chapter 05:
AIforce and partners
What AIforce and SIForce really are, and how Salesforce's new partner model shifts the interests of your implementation partner. - Chapter 06:
What to use when
Which AI tool fits which job, in twelve scenarios for manufacturers, from workshop prototype to ERP agent and EU data residency. - Chapter 07:
Costs and tokens
How token prices in development and Flex Credits at Agentforce runtime fit together, and how a manufacturer budgets for both. - Chapter 08:
AI code best practices
How to keep AI-generated Salesforce code under control with a rules file, pipeline checks and clear org boundaries. - Chapter 09:
Teamwork and delivery
How manufacturers set up roles, review gates, metrics and works council approval when partners and in-house teams deliver with coding agents. - Chapter 10:
Agents in manufacturing
Where Agentforce agents work for manufacturers today, how to build, test and secure them with rules, and what is only announced.
Articles
Articles in the masterclass, one question each
Each article answers exactly one question from day-to-day projects. Sorted by topic, so you only read what you need right now.
Agentforce in service, field service and sales 6 articles
When an agent pays off, which rules stay deterministic and how to test it before go-live.
- Agentforce for manufacturers: start now or wait?
- Agent Script: why business rules don't belong in the LLM
- Warranty Claims Adjudication Agent: what to prepare
- AI agents in Field Service: what works today?
- Testing Agentforce agents: when is one ready to go live?
- AIforce: Salesforce in Claude and Slack explained
Costs and licences 6 articles
Flex Credits, editions, Data 360 and AI coding: what it costs and how to estimate it upfront.
- Agentforce dealer service agent: cost per month
- Flex Credits: estimating Agentforce cost per action
- Salesforce editions Core, Advanced, Max for manufacturers
- MCP calls to Salesforce: a budget risk without a price
- AI coding costs in a Salesforce project: who pays?
- Data 360 costs for AI agents: catalogues and manuals
DevOps and quality 7 articles
How AI-generated code is reviewed, secured and taken through the pipeline.
- Reviewing AI-generated Salesforce code as the client
- API 67.0 user mode: why Apex sees fewer records
- Salesforce DX MCP Server end of life: what to change
- Quality gates for AI code in the Salesforce pipeline
- Agentforce Vibes, Claude Code, Codex or Copilot?
- AI rules file for Salesforce repos: who owns it?
- Vibe coding in Salesforce: prototype or production?
Partner governance and delivery 5 articles
Contracts, pricing models, team size and measurement when partners work with AI agents.
Integration and data 4 articles
MCP, hosted MCP servers and how agents reach SAP and older ERP systems.
Security and compliance 4 articles
Prompt injection, connected apps, works council, GDPR and the AI Act deadlines.
Contact
Planning an AI initiative on Salesforce?
A short conversation clarifies where to start and where not to.
Book a call (opens Microsoft Bookings in a new tab)Pick a slot in my calendar via Microsoft Bookings. Prefer to write? Use the form and I’ll get back to you.
Rather talk right away? +49 221 98656758