Dreamforce has wrapped up for another year. Hosted in San Francisco and streamed via Salesforce+, it was full of announcements, demos, and inspiring stories. If you’ve not had chance to watch Dreamforce, or want a quick re-cap, here are my take homes from Dreamforce 2023!
Artificial Intelligence (AI)
Salesforce has been working with AI since 2014. However, 2023 has seen a shift towards generative AI.
The largest AI event of the year
Dreamforce 2023 marketed itself as ‘the largest AI event of the year’. It pulled together various thought leaders from the world of AI. A few examples include:
- Dr Alondra Nelson: Harold F. Linder Professor, Institute for Advanced Study
- Ayanna Howard, PhD: Accomplished Roboticist, Entrepreneur & Educator
- Bob Sternfels: Global Managing Partner, McKinsey & Company
- Dario Amodei: Co-Founder & CEO, Anthropic
- Dr Fei-Fei Li: Sequoia Professor of Computer Science; Co-Director, Institute for Human-Centred AI, Stanford University
- James Manyika: SVP, Research, Technology & Society, Google
- Sam Altman: CEO, OpenAI
- Thomas Kurian: CEO, Google Cloud
- Will.i.am: Founder & CEO, FYI.AI
AI: Prepare, be careful and be optimistic
- Trust is key. Reassure users of data security (i.e., ownership and retention) and prevent misuse.
- Protecting data and intellectual property is particularly important for business.
- Ensure outputs do not cause more harm than good (e.g., avoiding toxicity and hallucinations/confabulations)!
- Opportunities and risks:
- We are in a period of rapid change (environmental, geopolitical, societal, and technological).
- Generative AI will have a large impact. For example, a study of 63 generative AI use cases highlighted that it could contribute between $2.6-4.4 trillion annually to the global economy.
- AI can help and harm. To maximise the benefits, be aware of and manage the risks.
- Plan now and re-evaluate:
- Current Large Language Models (LLMs) (e.g., GPT-4, Claude, PaLM 2, etc) are not the end. The technology is only going to improve.
- Individuals, businesses, and society need to adapt (e.g., reskilling, ensuring responsible use, empathy, representation, and governance).
With suitable safeguards, generative AI is an enabler. Responsible AI can empower individuals, teams or entire businesses to do more. The potential could be transformative.
Be sure to check out the penguin fact at the end for more insights on how AI can be a force for good!
It wouldn’t be Dreamforce without announcements! Here is my pick of the key updates.
Data + AI + CRM + Trust
Dreamforce 2022 focused on Salesforce Genie; harmonising data from across the enterprise. Genie has been renamed Salesforce Data Cloud and is processing over 2 trillion records per month. Since 2022, attention has moved onto generative AI and LLMs.
Businesses wishing to use AI need a robust data strategy. After all, if you use low quality ingredients in a recipe, the result may suffer. The same goes for data quality. For example, if a LLM uses outdated, incomplete or inaccurate data, the results will be no better. Salesforce Data Cloud, alongside the Einstein Trust Layer, provides tools to help manage these concerns. In doing so, it helps to power the next generation of CRM tools.
Salesforce is making this architecture open. New partnerships with Databricks and Amazon Redshift are making it easier to bring existing data lakes into your Salesforce tech stack.
Einstein 1 Platform
The biggest announcement from Dreamforce was the Einstein 1 platform. This is more than just a rebranding. Salesforce is unifying all products, including acquisitions, into the same core metadata framework. It will provide a single platform which is: “integrated, intelligent, automated, low/no code and open”. This all sits on top of Hyperforce: Salesforce’s next-generation infrastructure architecture, running on Amazon Web Services (AWS).
Key themes include:
- Marketing & Commerce Clouds: integrated into the Einstein 1 platform, with several updates over the next two quarters (see slides below)
- Einstein Trust Layer: benefits from generative AI, whilst protecting your business’s data (e.g., masking and ensuring zero retention in the LLM) and your team (detecting toxicity, and audit trails). For a deep dive into the Trust Layer, check out this video.
- Open and extensible: it is available to AppExchange and technology partners, simplifying future innovation. You can also bring your own model, such as Amazon Bedrock, SageMaker, Anthropic, Databricks, Vertex AI and OpenAI.
- Orchestration: use Flow to orchestrate and manage automation across the Einstein 1 platform; including Data Cloud Triggered-Flows! Check out Einstein co-pilot studio for more information.
Free Data Cloud & Tableau
Data Cloud can help unlock the value of data across a business from within Salesforce, without additional ETL tools. Coupled with Tableau, data analysis and insights become easier.
During the Dreamforce keynote it was announced that eligible customers will receive ‘free’ Data Cloud and Tableau licensing:
- Free licensing will be available from the 19th of September 2023
- Available to Enterprise, Unlimited and Unlimited+ edition customers. Self-enabled.
- Each account is given 250,000 data service credits (equating to ~10k profiles)
- Two Tableau Creator licenses are provided.
This provides an excellent opportunity to get hands-on with Data Cloud and Tableau. For most orgs this will only be enough to start exploring. It is unlikely to suit a Production use case, unless your org has very small data volumes!
Einstein copilot is a new generative AI tool. Users interact via natural language from within Salesforce. Common tasks can then be carried out and allow users to focus on more valuable activities. Copilot will be in every cloud.
For more information on copilot, check out this article.
Einstein Copilot Studio
Einstein Copilot Studio is a no/low code solution to provide your Copilots the skills they need to succeed. Much like adding a skill to Alexa, add skills to your Copilots to suit your use case.
- Prompt Builder: Create reusable prompts which are dynamically grounded. Pilots of Email and Field Generation are available in October 2023.
- Model Builder: Either use a LLM created by Salesforce, or deploy and train your own models using CRM data.
- Skills Builder: Create and add custom AI-skills for a Copilot to complete.
In addition to Einstein Copilot Studio, Salesforce is working on a number of tools to help admins and developers. Some of the features which caught my attention were:
- Einstein for Flow. Create a draft flow with a prompt from Flow Builder.
- Flow Repeaters. Ability to repeat logic within a Flow (on a single screen).
- Scale Centre. Self-service, near real-time analytics of your org. Helps identify potential performance and scalability issues. Available in Unlimited Edition.
- Code Analyzer Extension. Visual Studio Code Extension. Scans code and produces a list of violations requiring attention to improve code quality. For more information, click here.
- Einstein for Developers. Create Apex code with a prompt from Visual Studio Code.
- Code Builder. Modern, web-based integrated development environment. This is becoming generally available in October.
If you want to see pitches from three inspiring entrepreneurs using AI to make a difference, check out Dreampitch. With $250,000 USD investment on the line, who will win? Don’t worry, no spoilers about the winner here!
- Wysa: A mental health app using generative AI to help support those in need. Checkout the app on Google Play / Apple Store (the mascot is a penguin, so extra credit here!)
- Treeswift: Helping foresters manage forests around the world. By scanning under tree canopies and AI to analyse the results, Treeswift provides detailed insights. Essential for maintaining biodiversity, preventing wildfires, and the fight against climate change.
- Ellipsis Earth: Using machine learning to rapidly survey and gain insight into pollution (e.g. mapping sources of marine pollution). This helps inform appropriate policies.
True to the Core
True to the Core is a forum where community members ask questions and share ideas with the Salesforce Product Team. Some of the agenda items included:
- Dynamic Forms. Continuing to improve functionality (e.g., visibility controls, performance).
- List Views. Migration from Aura to Lightning Web Components (LWC) continuing.
- OmniStudio vs Flows. There has been blurring between these two products due to product teams working together. OmniStudio will be receiving dedicated focus soon.
- Setup. Common pain points discussed (e.g., inconsistent UI and slowness). Salesforce is committing resource to improve setup.
- Reporting. Core Salesforce Reporting and Dashboards are still being enhanced (e.g. Winter ’24 supporting data sourced from Data Cloud).
- Scale centre. Provides information on how to improve Report performance. It was suggested that these insights are built into Report Builder itself to empower users.
- SKUs (Stock-keeping units). Salesforce is looking to simplify SKUs and move into ‘bundled’ packages (e.g., Unlimited+)
- Campaign and Campaign Member. Enhancements may be on the way to improve reporting and connectivity to other areas of the platform.
Skills: Future-proofing your career
Salesforce clearly views Data Cloud and Einstein’s Generative AI capabilities as a potential gamechanger for its customers. It also recognises the need for Salesforce Professionals to gain new skills to help its customers gain the best from these technologies.
Just before Dreamforce, two new certifications were released:
Alongside this, there were a number of great sessions covering how to upskill. Here are some of my top picks:
Dreamforce may be over for another year, but the possibilities for innovation will last much longer. Without doubt, Salesforce is focusing on generative AI. The content in this post has explored many of the features which are either currently, or will soon be available. Time will tell how effective each feature is. However, the direction of travel is clear and the potential gains significant. It will be interesting to see how things progress between now and Dreamforce 2024!
Until next time, thanks for reading!
Bonus Penguin Fact
As Dreamforce has been very AI focused, it is only fitting the penguin fact is too!
Next time you visit a hospital, did you know a robotic penguin may be on hand to help you get discharged quickly? It’s true! Pioneering work at Milton Keynes hospital, United Kingdom, is being carried out. The autonomous robotic penguins, called Milton, have been created by the Academy of Robotics. They are being trialled to help free beds faster by delivering medicines to patients ready to leave. Click here for more information.
Found this post useful? Why not share!
Or check out some of our other articles: