AI Demo Jam promotional image

Please note: Alfred is now called Vinton but this blog was published prior to the name change.

On Thursday 15th February we participated in Salesforce Ben’s very first ever AI-related Demo Jam. Salesforce Ben invited eight Salesforce ISVs, who are using AI to transform the way we work in Salesforce, to spend three minutes each demoing their solution online to an audience of over 1000 attendees!

Let’s run through the different solutions that were presented. They’re all a testament to the innovation to be found in the Salesforce ecosystem.

Elements.Cloud

The first up for the Demo Jam was Elements.Cloud. Elements is a change intelligence platform that helps accelerate your time to value from Salesforce. It allows you to make intelligent change decisions based on a shared understanding of org configuration, dependency analysis and documentation. Powered by AI, it dramatically shortens the time to draw process diagrams write complete user stories, and create solution recommendations.

Ollie Jeffcote from Elements.Cloud then presented the AI features that help you accelerate the way you do business analysis. First he generated a process map using an example related to the demo jam, demonstrating the versatility of how the tool can generate pretty much any process. Ollie created a new map about “How to do a Demo Jam” and entered a prompt for this. He then left this running in the background.

Elements.Cloud example screenshot

Next Ollie viewed an existing process map for an IT help desk. The scenario he gave was that the demo company needs to make some changes to Salesforce based on this process, so needs to generate some user stories. Using Elements you can generate user stories using AI in just a few clicks. The brilliant thing about this is that those stories are going to be relevant to the business process, going to take into account the outcomes that you’re expecting, the systems you're using, and the people involved in those processes as well. Generating the stories this way is great for reducing the chances of having technical debt down the line.

Elements.Cloud example screenshot

Ollie showed an example user story with a description, and five bullet points of acceptance criteria, all generated using AI based, on the process that Ollie had drawn out. Once you have the story, you need to think about how to implement it and Elements can even help with that. Clicking the Ask Elements button will suggest the objects, the fields, and the permissions to use to support and implement that user story. In under three minutes, Ollie managed to get from a process diagram to the implementation.

  • Elements.Cloud example screenshot
  • Elements.Cloud example screenshot

Ollie then returned to the original process he’d asked Elements to create with a prompt about “how to do a demo jam.” The full process was mapped out.

Ollie’s ElementsGPT demo illustrated the incredible speed that you can go from an idea of a process through to understanding how that needs to be implemented.

Elements.Cloud example screenshot

Metazoa

Bill Appleton from Metazoa was up next to demo Metadata Studio.

Metazoa builds a product called Snapshot, and Snapshot makes it easy to manage your Org for administrators and developers. With tools for documentation and compliance, optimization and cleanup, and security and permissions. Bill's demo focused on Metadata Studio, a Prompt Engineering Platform. With 100+ pre-built prompts and customisable designs, the platform’s prompt automation features enable seamless transformation of org data and metadata. Metadata Studio removes technical debt, generates documentation and optimises security with AI effortlessly.

Let’s see what Bill demoed.

Metadata Studio is a prompt engineering platform that solves very practical problems for Salesforce administrators. Easy to use with 100 pre-built prompts, you can just click on any of them and click the submit button to solve your problem. You can also switch over to prompt engineering to build your own prompts and share those with your team.

Bill gave a really practical example of how this might be used. Say you're a Salesforce Administrator getting started in a new Org, and there's a hundred flows for which you have no idea what they do. Bill demoed clicking a prompt called “Create Flow Node Descriptions”. He ran the prompt and the selected a flow and the AI adds descriptions to the metadata which you can validate and save into your Org.

  • Metadata Studio example screenshot
  • Metadata Studio example screenshot
  • Metazoa Studio for Health Demo screenshot

Bill then showed what that looks like when you’re viewing that same Flow in Salesforce. At ever stage of the flow, each box has a description and you know exactly what they do.

Metazoa example screenshot of Salesforce flow

Bill then showed that you can apply the same prompt to any number of selected flows, not just one at a time. You can also create a developer project for your release management team to move all of these changes into production.

Administrators have always been responsible for data quality and Bill showed a couple of other prompts that will help with data quality. One creates new descriptions for every account and opportunity in your org. Another will create account contact roles for all of your accounts or opportunities. Taken together it revolutionizes the data quality that your salespeople are going to be working with.

Doing the same thing for developers, Bill showed how you can also create an Apex class dictionary. The class dictionary has HTML output and when you build that, you see that all of your Apex classes are fully documented including their external Integrations.

Alfred (now Vinton) by Nativevideo

Next up was our very own Luca Benini presenting Alfred (now known as Vinton) . We had a great intro from Salesforce Ben who kindly mentioned that his team is a proud customer of Nativevideo and described Alfred as a “great, great product.” Thanks Ben for the shout out!

A quick intro to Alfred. It’s an AI meeting assistant that joins calls to transcribe, summarize, suggest next steps and draft a follow-up email, all captured in Salesforce against the relevant contact, account, opportunity or lead. Alfred also comes with audio note functionality so you can record voice updates after customer interactions and Alfred does the rest. Alfred improves sales productivity and the quality and quantity of customer information in your Org.

Luca started the demo in the context of an Opportunity. Luca demoed the audio note functionality, leaving a note following a conversation with a prospect at this particular opportunity. Recording your updates in a conversational way is very natural. Luca left that audio note to process whilst showing other features of Alfred.

Screenshot of an example opportunity in Salesforce with an audio note being recorded within it

Luca switched over to a Contact to show you what a video call looks like using Alfred. He showed the example of call captured in Salesforce against the contact, with the transcript available, a summary generated by the AI, a follow-up email that can be copied into your email client, and suggested next steps that can be easily (with one click) be turned into a Salesforce.

Salesforce Screenshot with an example presentation and summary from a video meeting

Returning to the earlier audio note recording, Luca showed how the AI had captured the gist of what he was saying in his note and how the AI had once again generated the summary, next steps and email follow-up.

Salesforce Screenshot of an example summary of an audio note recording

After our demo we received a couple of great questions about Alfred which we’ve answered here:

1. How does Alfred cope with multiple languages?

Alfred handles multi-lingual speech-to-text, transcription, and translation into 30+ languages and accents. Some customers will transcribe recordings in their local language, but the call summaries will be added in English.

2. How does Alfred differ from Einstein Conversation Insights?

  • Alfred has voice note updates as well as the call recordings, so users can record ad hoc voice updates when required
  • Alfred can be used in any Salesforce cloud - not just Sales Cloud. We have customers using Alfred with Service Cloud or for recruitment with a Salesforce-compatible ATS
  • Alfred has flexible AI prompts that are completely customisable
  • You can include Alfred in flows, so you could drop in the Alfred audio recorder as part of a flow
  • If you need to share a call recording with someone outside of your Org, you can generate a sharing link and share the recording, summary, next steps, and the transcript, visible on a custom link

Pipelaunch

Next up was Pipelaunch.

Pipelaunch adds a real-time view of your customers by integrating with 20+ data sources. It helps you:

  • find the ideal customer profile based on matching LinkedIn profiles without leaving Salesforce
  • engage with your prospects using GPT powered sales messages
  • keeps your Salesforce data automatically updated

Donny from Pipelaunch explained how Pipelaunch is all about helping Sales reps to love Salesforce. He showed an example of a new lead inbound from the website. Viewing the lead in Salesforce, the details for the lead were matched to a contact in LinkedIn - which shows in the Pipelaunch component to the right of the lead details. The sales person can try to match the LinkedIn profile to the lead in Salesforce, getting the profile picture, the title, the country location. There’s also much more data to view such as their experience but also information about the company where the prospect is actually working.

  • Pipelaunch example screenshot of LinkedIn integration and contact and company insights
  • Pipelaunch screenshot example

Donny then showed how AI can help in reaching out to the prospect. The record had no email address when it came into Salesforce, but it can be enriched by Pipelaunch and an email address can be validated and added to the record.

Using AI, Pipelaunch can help create a cold email message for the prospect. You can add extra information before generating the email to help the prompt, such as selecting the product you want to talk to the prospect about.

Pipelaunch screenshot of an example contact

Pipelaunch also has a Chrome Extension. If you visit the company website for a prospect you are targeting, Pipelaunch leverages AI to extract information relevant for this company.

Rocketphone.ai

Rocketphone came next with a very impressive demo. Apart from our Alfred demo, I must say that this one was my favourite.

RocketCell is the business mobile network that records, transcribes, summarizes and logs every mobile call in Salesforce. No app is required!

Muj and Sam from Rocketphone pretended to be prospect and sales person discussing a broadband offer. The call was made natively on the salesperson’s mobile phone, directly from one phone to another. Halfway through the call they asked Ben to give a topic for them to discuss to prove the call was live. Muj highlighted that Rocketphone is the only company in the world that has native Network to Salesforce integration.

As soon as the call ended, the salesperson received two WhatsApp alerts from ZanderGPT. ZanderGPT is Rocketphone’s GPT chat bot which figured out that the customer wants the order form, the terms and conditions, so sent a WhatsApp to the salesperson which said, “Just say yes and I'll send the documents over to you.” By saying yes, the documents are sent to the sales person to review!

Rocketphone demo screenshot

Muj then showed his screen and his Salesforce Org where you can see in the activities, the call that occurred, the transcript with the recording itself which you can play back if you want to. There’s also a call summary and coaching tips too!

Congratulations to Rocketphone for winning the "Most Creative Demo" at the Demo Jam. 👏

Precursive

Precursive is the service delivery Cloud for Salesforce helping Professional Services teams forecast and execute customer projects. Precursive demoed building a project from a prompt using ChatGPT to suggest project risk mitigations and best-fit assignment of resources based
on availability and skills.

Andy from Precursive started the demo by giving the scenario of trying to scope a project, but not always having the information you need. Precursive project composer, powered by AI, will help you to scope and forecast an entire project from a short prompt. This process can take around a minute. The prompt was pre-filled from the opportunity description, but it could equally be pulled in from meeting transcriptions from Gong, Zoom Teams or even Alfred or Rocketphone.

Precursive project composer screenshot

Precursive takes the prompt provided by the user and grounds that with data from Salesforce and Precursive, including account, opportunity, your own project methodology, roles and rates, you can describe your delivery methodology in natural language and Precursive will use this to build projects for you.

The AI will also be looking to maximize budget, adjust the rates on participants to maximize the margin and schedule the Project based on the information that you're providing.

Andy showed how the project was built in less than a minute using AI, with AI-generated tasks and a Gantt chart with full dependencies automatically scheduled. From the kanban you can see how the actual roles are assigned to each task. These tasks and roles all contribute to a resource forecast within Precursive, so this means that this project is now going to contribute to this overall view of resource Demand by role for the demo organization. There are also AI
generated risks and assumptions.

Precursive gantt chart screenshot

Andy also showed the “Risk Whisperer” which provides some risk mitigation. It takes some information about the risk and suggests a way that you can mitigate this problem. Once the project is set up, you can assign people. Precursive’s Magic Assign recommends people who are available and have the right skill to pick up the project.

Precursive in Salesforce screenshot

Provar

Provar improves the predictability speed and quality of your releases for Salesforce and beyond. Its low code solutions provide intuitive and resilient test automation, integrated quality management and outstanding support emowering teams at every stage of the development life cycle. Provar helps teams improve release agility and reduce system defects, so you can spend less time testing and more time innovating.

Will Hughes from Provar ran the demo. He gave a scenario about a call from a customer who had a huge backlog of user stories that he needed to write tests for, due that evening so time was tight.

Provar has been developing AI capabilities to help save time with repetitive tasks. Will showed that if you import the Jira project you’re working on, you can then import every user story or issue that you need to write a test for. Because there is a description and acceptance criteria available, Provar has feature called “Generate test cases” which means that at the click of a button they can use AI to scan to scan the description and acceptance criteria you've got in your user stories and then will develop five or six test cases that you can go and review and choose from.

Provar screenshot example

Tim then gave another scenario where they need to create a report based on bugs the customer is seeing from their Apex test. Provar has also built an AI-powered capability to help with complex and analytical tasks just like this scenario. Going to the test plan in Provar, the test cycle where the failure is occurring will show the defect that’s been automatically raised by Provar. Instead of going to that defect and doing research and analysis, now you can generate a root cause analysis report at the click of a button which will give a full description of what the issue is, a breakdown of what the root cause of that issue is, and also recommendations with how to improve that. You can hand that report directly back to your developers.

Finally Will showed a new Provar Assistant (currently in beta) which is an AI powered chatbot to answer any question you have which is related to Provar.

Swantide

Swantide is the first AI-powered application on the AppExchange that helps companies implement manage and scale their Salesforce instances. For organizations just getting started on Salesforce, Swantide can cut down implementation from months to minutes with its AI powered library of pre-built customizable templates, while significantly reducing cost. Swantide personalises your Salesforce for intelligent debugging and question answering.

Taylor Lint, founder and CEO of Swantide shared just some of the capabilities that Swantide has for admins, consultants and developers.

Taylor demoed as if she was an Admin who had received an urgent email from someone on the team saying that they thought a lead should have been created but it wasn’t. It was related to a specific custom object in the system. By using the AI assistant in Swantide, the Admin is able to ask a question about why the issue might have occurred and the AI provides:

  • a little bit about the flow
  • what could be happening
  • how to investigate further
Swantide example screenshot

When you connect to Swantide, it’s automatically personalized to your Salesforce instance. Swantide takes care of your metadata, using AI to document all of that metadata. It starts keeping track of all the changes to your metadata, can connect to your ticket backlog that your Admin team has (so if you ticket history in Jira or Zendesk or ServiceCloud for your admin team) to further personalize your AI assistant.

You can also upload documents that are related to your sales process or customer service process. You can ask the AI assistant anything about how your system is set up, changes that might have been made, and it can help you intelligently start to debug things and suggest next steps.

Swantide's data dictionary allows you to pull up a piece of metadata and you can see the documentation and implementation details completely generated by AI.

Swantide example screenshot with metadata components

The AI assistant is also available for your Salesforce end users to use, so if they are constantly asking common questions about how to do things specific to your Org, they can ask the AI assistant rather than coming to the Admin team.

Congratulations to Swantide for winning the “Best App for Everyday Use.” 👏

Phew! That was one heck of an AI Demo Jam and a lot of new information to consume. We highly recommend spending time watching each of the demos and go ahead and try out each of the solutions in your Org, especially if they come with a free trial or freemium offer like Alfred does!

We’re looking forward to participating in future Demo Jams in the Salesforce community and seeing how AI evolves over the next year.

You might also be interested in