Each year, TrailblazerDX brings developers, admins, and architects together to discover the latest innovations across Salesforce. This year’s event put the “AI” in “Tr-AI-lblazer” with hands-on workshops, dozens of demos, and hundreds of technical sessions that showcased how to use the latest innovations in generative AI on the Salesforce Platform. We launched new products for developers, including Einstein 1 Studio, Data Cloud Vector Database (in pilot), and new AI-powered developer tooling on Einstein for Developers.

In this blog post, we’ll give you all the highlights from TDX ’24 to help you enhance and augment your workflows with AI. You can also find the recorded TDX ’24 sessions on the Salesforce Developers YouTube channel in these two playlists:

AI Sessions for Developers | TrailblazerDX 2024
Developer Sessions | TrailblazerDX 2024

Einstein 1 Studio

Salesforce launched Einstein 1 Studio, the new unified platform for integrating customer data and generative AI in a security-first environment across Salesforce. You can use Einstein 1 Studio’s Prompt Builder, Copilot Builder, and Model Builder to build AI-enhanced applications and workflows.

Prompt Builder (GA): With Prompt Builder, you can create prompt templates that ground your AI prompts with data through CRM merge fields, external data with Data Cloud, and other services with Apex to give LLMs the data they need to give you the answers you want. You can use Prompt Builder templates in Lightning, Flows, and REST to populate fields and generate content, such as sales emails, case summaries, reply recommendations, and more.

Copilot Builder (Beta): Using Copilot Builder, you can build custom Copilot Actions using your own custom Apex classes, Prompt Builder templates, and Flows. Define the action instructions, outputs, and any required inputs, so that Copilot knows how and when to use a Copilot action. Once you’ve defined an action, you can assign it to your Copilot, so that it can be used throughout Salesforce.

Model Builder (Beta): With Model Builder, you can create and train predictive models on your data in Data Cloud using clicks instead of code. You can also connect to models you’ve created in Amazon SageMaker, Databricks, and Google Vertex AI. Use your custom-built models to predict outcomes by creating prediction jobs in Data Cloud or Flows.

You can check out each of these new tools by watching their recorded talks on Salesforce+:

• Get Started with Prompt Templates for Developers (TDX ’24)
• Get Started with Einstein Copilot Actions for Developers (TDX ‘24)
• Bring Your Own Model with Model Builder (TDX ’24)

Data Cloud Vector Database

Generative AI works best when it has access to relevant information. With Data Cloud Vector Database (Pilot), you can set up a search index to retrieve knowledge from structured and unstructured data sources, such as knowledge articles, PDFs, and call transcripts, using clicks instead of code.

Use Data Cloud Vector Database to generate a semantic search index from your data model objects. Once you’ve created your search index, you can retrieve structured and unstructured data in your Apex code (and by extension, Prompt Builder templates), flows, and analytics.

Einstein for Developers

The Einstein for Developers VS Code extension released several Beta features to help you write better code faster when building Apex code and Lightning Web Components, including inline autocomplete, test generation, and natural language to code.

Natural language to code (Beta): We’ve introduced a sidebar that you can use to generate Apex and Lightning web components with natural language. Describe the code or component you want to generate, and let Einstein write the code for you. Referencing your code in other files? Open the files in a tab (focus or otherwise), and Einstein will be able to reference them, too. You can also ask Einstein to generate your questions from the Command Palette.

Inline autocomplete for Apex and Lightning Web Components (Beta): Now, you’ll receive generative AI-enhanced inline autocomplete suggestions as you write Apex code and build Lightning web components. Hit the tab key to accept the suggestions and speed up your coding without breaking your flow.

Test generation for Apex (Beta): Right-click within an Apex class file and select “Generate Test” to have Einstein create a test and improve your overall code coverage.

You can start using Einstein for Developers today by downloading the extension from the Visual Studio Code Marketplace. Be sure to check out the documentation for details on how to best use all of Einstein’s features.

TDX trailmixes

TDX ’24 was jam-packed with hands-on learning sessions, demos, and workshops. Check out the TDX trailmixes below to learn more about our latest AI innovations and platform best practices.

Data Cloud
Explore Data Cloud
Data Cloud Activations
Data Cloud Ingestion

Generative AI
Get Started with Prompts and Prompt Builder
Explore Einstein Copilot
Einstein for Developers

APIs and Development
Salesforce APIs
Apex
Lightning Web Components
Platform Events & Change Data Capture

Developer Tools
Salesforce CLI
Salesforce IDEs
DevOps Center

Automation
Flow Builder
Flow Orchestration

Low-code & Configuration
Lightning Design System
Lightning Experience Builder & CMS
Lightning App Builder

Security
Privacy Center & Data Mask
Shield & Security Center
Permissions & Security Controls

Scale and Integration
AWS Unified Developer Experience
Sandboxes on Hyperforce

Other
Salesforce Mobile
Salesforce Backup

Conclusion

TDX ’24 was packed with releases, workshops, and talks that put generative AI and data at the core of the platform. You can build AI-powered workflows with the Einstein 1 Platform, use semantic search through Data Cloud, and write code faster with natural language and smart suggestions using Einstein for Developers. Check out the sessions that deep-dive on these topics and more on the developer highlights track from TDX ’24 on Salesforce+.

Want more AI content?

Learn how Einstein 1 Studio, including Prompt Builder, Copilot Builder, and Model Builder, allows you to develop custom AI solutions, grounded in your customer data, with any model you choose to customize your Einstein Copilot and extend AI for CRM with low-code and no-code tools.

Join live or watch replays of each episode of the Developers AMA: Einstein 1 Studio series to see the latest product demos, and get your technical questions answered directly by Salesforce product experts in real time.

Wednesday, March 27, 9 a.m. PT, Prompt Builder
Want to maximize productivity with AI-enhanced custom prompts?

Join host Principal Developer Advocate Alba Rivas as she dives into Prompt Builder alongside product managers Joshua Schneyer and Avanthika Ramesh to explore how to build, ground, and invoke prompts that connect to your CRM data via Apex or Flow for secure and personalized responses.

Wednesday, April 24th, 9 a.m. PT, Copilot Builder
Ready to build standard and custom actions to extend Einstein Copilot?

Join host Lead Developer Advocate Aditya Topalli as he explores Copilot Builder with VP of Product Management Vinay Balusubramaniam and Principal Architect Prithvi Padmanabhan in a live Q&A conversation. Learn how to build standard and custom actions to extend and customize your copilot using Apex, Flows, or prompts to integrate familiar platform features.

Wednesday, May 29th, 9 a.m. PT, Model Builder
Did you know you can build or bring your own AI models, including LLMs, to Salesforce?

Join host Developer Advocate Danielle Larregui for a live Q&A with Executive VP of Software Engineering, Raveendrnathan Loganathan and Senior VP of Product Management, Gabrielle Tao. Explore how to develop custom AI solutions, grounded in customer data, with Model Builder and Data Cloud.

About the author

Charles Watkins is a Lead Developer Advocate at Salesforce and a full-stack software developer focused on the core Salesforce Platform. You can find him on X.