No Jitter Roll: AWS Climbs to New Heights with Gen AI Updates, Zoom Automates Workflows, Sennheiser Debuts Zoom Room Certified Bars and Avaya's CEO to RetireNo Jitter Roll: AWS Climbs to New Heights with Gen AI Updates, Zoom Automates Workflows, Sennheiser Debuts Zoom Room Certified Bars and Avaya's CEO to Retire
At its New York Summit AWS announces multiple Gen AI updates, Zoom launches a workflow builder while Sennheiser launches Zoom-certified TeamConnect Bars, Egnyte debuts Gen AI Copilot, SuccessKPI and ConnectGen launch Contact Center Builder.
July 12, 2024
Welcome to this week’s No Jitter Roll, our regular roundup of product news in the communication and collaboration spaces. Leading off this week, we highlight: Multiple Gen AI products and updates from AWS, a Contact Center Builder offer from SuccessKPI, the new Gen AI-powered Egnyte Copilot, Zoom’s new workflow builder tool. and the Zoom Room-certified TeamConnect Bars from Sennheiser. Lastly, Avaya announced that Alan Masarek, its President and CEO, will retire at the end of the calendar year.
Also, we checked in with analyst Derek Top of Opus Research for some context on two of AWS’s announcements.
Gen AI News from AWS
From the AWS Summit New York, AWS made several Gen AI-related announcements including the general availability of Amazon Q Apps, AWS App Studio (in preview), new Amazon Bedrock features and a partnership with Scale AI. The following illustrates what AWS terms its generative AI stack – applications at the top, the tools with those apps can be built in the middle to the infrastructure layer at the bottom.
Amazon Q Apps
A feature of Amazon Q Business, Amazon Q Apps (now generally available) allows users to describe the application they want in a prompt and Amazon Q then generates it. Amazon Q also gives employees the option to generate an app from an existing conversation with a single click. AWS said that during the preview of Q Apps, it “saw users generate applications for… summarizing feedback, creating onboarding plans, writing copy, [and] drafting memos.” This blog post describes how Q Apps Creator works.
AWS App Studio (in preview)
AWS App Studio allows a user (e.g., IT project managers, data engineers, and enterprise architects) to write a basic prompt (using natural language) that describes the application they want. App Studio will generate an outline to verify the user’s intent and then build an application with a multi-page UI, a data model, and business logic. Users can then iterate on the generated app by interacting with App Studio. Examples of use cases include claims processing, inventory management, project approval, etc.
The application can be connected to AWS systems (e.g., Amazon Aurora, Amazon DynamoDB, and Amazon S3), Salesforce, and other third-party services (such as HubSpot, Twilio, and Zendesk) using an API connector.
Fine-Tuning and More in Amazon Bedrock
Amazon Bedrock provides customers with access to multiple large language models (LLMs) as well as tools to build Gen AI-powered applications. AWS customers can now fine tune models in Bedrock by providing labeled data in Amazon Simple Storage Service (Amazon S3). AWS further announced that Bedrock supports more data sources for retrieval augmented generation (RAG) including document repositories, databases, and APIs. AWS also expanded the list to include connectors for Salesforce, Confluence, and SharePoint (in preview).
Additionally, AWS added vector search to some of its data services, including OpenSearch Service and OpenSearch Serverless, Aurora, Amazon Relational Database Service (Amazon RDS) and Amazon MemoryDB.
(Editor’s note: Fine-tuning creates a copy of the model adapted for specific tasks; RAG lets existing models reference new datasets when generating responses. Vector search uses mathematical vectors to search unstructured data.)
Agents for Amazon Bedrock
Agents for Amazon Bedrock is an existing capability. What’s new is that these AI agents can now retain state across multiple interactions to remember where a user last left off and thus provide better recommendations based on prior interactions. For example, a flight-booking AI agent could reference the last time a customer traveled. Also, AI Agents can now interpret code to handle use cases, such as data analysis, data visualization, text processing, solving equations, and optimization problems. This blog post covers both memory retention and code interpretation in more detail.
(Editor’s note: An AI agent is a “software program that can interact with its environment, collect data, and use the data to perform self-determined tasks to meet predetermined goals” set by people. Emphasis ours.)
Guardrails for Amazon Bedrock
Guardrails is another existing capability within Bedrock. What’s new is that AWS has added contextual grounding checks to detect hallucinations in model responses for applications using RAG and summarization applications.
Ask an Analyst: Derek Top, Opus Research
No Jitter asked Derek Top, Research Director & Senior Analyst with Opus Research, to provide some context on these two aspects (fine-tuning and AI Agents) of the AWS portfolio
NJ: How do these features benefit end users?
Derek Top: There is a lot here for companies that want to roll up their sleeves and build applications with Gen AI, even experiment with creating their own models and fine-tuning LLMs. Putting more capabilities into the hands of developers — and business users, for that matter — is an incremental (and necessary) step in the integration of Gen AI into employee workflows and business processes which ultimately results in end user benefits.
From a CX and contact center perspective, AWS is playing a long game where they anticipate both agents and customers are going to get more conversational in their input to chats, voice bots and the dialog slots that are starting to appear everywhere.
(Editor’s note: The term ‘dialog slots’ refers to “slot filling, which is the process of collecting required user input.”)
NJ: How do these features match up to the other LLM providers? Is Amazon playing catch-up? Breaking new ground?
Derek Top: AWS isn’t necessarily breaking new ground so much as building on its solid foundation. They have created an environment in the cloud to make it “safer” for enterprises to merge their own models with foundation LLMs so that they can use RAG to offer answer questions from customers and agents alike.
For enterprises, the big challenge of integrating Gen AI and LLMs into existing workflows is encapsulated in the word “de-risking.” Bedrock is ready-made for that, and AWS is a trusted brand that has proven that it works well with others (including other hyperscalers) while it pursues strategies to build its own LLMs and Gen AI resources.
SuccessKPI and ConnectGen Launch Contact Center Builder
The cloud-native, contact center and CX platform provider, SuccessKPI, partnered with ConnectGen to launch Contact Center Builder, a CCaaS solution for small contact centers. The Contact Center Builder (CCB) guides the user through a browser-based setup wizard. CCB then completes the setup of an interactive voice response system (IVR) and interactive virtual agent (IVA) chat bot. It also provides automatic call distribution (ACD) for directing calls to human agents, and an instance of Amazon Connect for handling calls, reporting and measuring agent performance. The solution can also integrate with CRM systems.
Zoom Launches Workflow Automation Tool Beta
The collaborative workspace company introduced a workflow builder into Zoom Workplace. Users with the feature can craft a workflow via a drag-and-drop interface and automate communications workflows like introducing new members to a channel, making recurring reminders for project status updates, and submitting time off requests. Users can set up sequential workflows, conditional workflows based on if-then logic, or switch-based workflows that automate multiple paths. In the latter case, that might look like someone setting up automated routing in a channel to move user requests to billing, IT support or HR, based on keywords in the initial user request.
“We built Workflow Automation to be easy for teams of all sizes and abilities to use. We’re launching Workflow Automation with Team Chat first because it’s an opportunity to strengthen collaboration with team members and get work done asynchronously,” said Wei Li, head of Zoom Team Chat at Zoom. “Workflow Automation helps teams by taking the guesswork out of setting up workflows and helps cut down on tedious and repetitive tasks.”
During the beta period, Workflow Automation is available to paid Zoom users. Once the feature is generally available, however, workflows will be classified as standard or premium, and paid plans will include unlimited standard workflow runs and allotments of premium workflow runs.
Egnyte Launches Generative AI-based Copilot
The content collaboration and governance platform provider announced the general availability of Egnyte Copilot. New and existing customers on eligible plans can designate specific groupings of files as “knowledge bases” and configure each base using domain-specific prompt libraries to guide the AI’s responses. These “bases” can be created by non-admin users around: projects (clients, marketing campaigns), file types (contracts, presentations, resumes, etc.), and knowledge domains (product specs, safety regs, HR benefits, etc.). It appears that the Copilot will generate answers from the information in those “bases.”
Additionally, Egnyte Copilot can extract and describe specific information within images, summarize the contents of audio and video files, create transcripts, and surface insights and citations. Users can prompt the model with context-specific instructions.
Egnyte Copilot can be accessed via Egnyte’s web, mobile, tablet, and desktop applications.
Sennheiser TeamConnect Bars Now Certified for Zoom Rooms
Sennheiser announced that its TeamConnect Bars (TC Bars) are now Certified for Zoom Rooms. This certification enables Zoom users to conduct hybrid meetings with Sennheiser audio and 4K Ultra HD video. The TC Bars are available in two models:
TeamConnect Bar S: Four microphones and two speakers for small rooms.
TeamConnect Bar M: Six microphones and four speakers for mid-sized rooms.
TC Bars are plug-and-play USB devices and feature an onboard Dante port (an A/V connectivity platform) for extension mics and an option to add an external USB camera. TC Bars support multiple mounting options and remote management via Sennheiser Control Cockpit.
Avaya News: CEO Leadership Transition
Avaya announced that Alan Masarek, its President and CEO, will retire at the end of the calendar year. His replacement will be Patrick Dennis, the current Chairman of the Board, effective September 1, 2024. For more, read Dave Michel’s article on this big Avaya news.
This Week on No Jitter
In case you missed them, here are some of our top stories:
Microsoft Copilot versus Apple Intelligence: Microsoft’s AI-powered digital assistant has made enterprise inroads. Apple’s new AI-powered features in Siri could add another copilot to users’ everyday computing experience.
An Eight Step CX Roadmap: An overwhelming array of AI possibilities for contact centers makes a structured evaluation approach imperative. Use this one as your starting point.
CX Doesn’t Matter: Many companies are not really asking themselves how to improve their customer service. Rather, they’re engaging in active experiments to figure out how much they can get away with.
Reserve Your Spot: Enterprise Connect AI 2024
With AI moving at a frenetic pace and IT leaders scrambling to keep up, this new 2-day event from Enterprise Connect will provide up-to-the-minute, in-depth, unbiased content with conference tracks covering CX, Productivity, and IT Management to help the Enterprise IT community leverage AI to advance the enterprise. Learn More about Enterprise Connect AI, October 1-2, 2024, Santa Clara, CA.
