AI & AWS: Making Work, Less Work

Xebia Background Header Wave

In our first post from AWS re:Invent 2023, we called this "the year of AI" – and we touched on several announcements by Adam Selipsky, regarding the AWS approach to AI & ML. The evening keynote – featuring AWS Applications VP, Dilip Kumar, and business leaders from Asana, U.S. Bank, and Woodside Energy – offered us “the next deepest level of detail,” including some significant news from the AWS Applications group.

Dilip’s focus was clearly on customer service, and (even) employee happiness, as well as how GenAI is improving business outcomes – for example, Amazon Connect, when empowered by Amazon Q. Connect is an omnichannel cloud contact center app that lets you create personalized experiences for your customers, offering either chat or voice contact, based on factors such as customer preference and estimated wait times. But with the added power of Amazon Q, which bundles AI & ML with Connect, the app can generate real-time recommendations for the agent’s “next best action” or "most relevant answer," and produces summarizations of calls, immediately after hanging up. In fact, the embedded GenAI is so quick and capable that it can generate "summaries of all the call summaries" for supervisor review, revealing previously unseen patterns across thousands of calls.

Amazon Connect

Bottom line, through GenAI, Amazon Q helps Amazon Connect support and improve both the front-end and the back-end of customer calls, saving money and improving call quality.

Another case springs from the axiom, "Disconnected apps hinder productivity." The context is familiar to every worker – they open six or seven different apps, in order to take on a particular task or project, then jump back-and-forth between these programs. But AWS App Fabric can leverage GenAI to generate insights from multiple SaaS applications, simultaneously and while maintaining the overall "work context." Meanwhile, GenAI works behind the scenes, surfacing relevant content and maintaining the work context. The practical benefits? Employees can complete everyday tasks faster using GenAI, and IT and security teams can more easily manage and secure applications using a standard schema.

AWS App Fabric

For our part, Xebia is also striving to "make work, less work" by leveraging AI to create a smarter way of working, with happier employees, more connectivity, and faster completion of everyday tasks. Our innovative software developers have recently announced the alpha version of Xebia AI Chat Assistant, running on AWS. Chat Assistant extends the reach of AI to private company documents and records, preserving confidentiality while empowering employees with the query-and-analysis capabilities associated with GPT 3 or 4. Further, Chat Assistant lets you choose your cloud, choose your LLMs, and choose the “role” through which you engage a particular use case.

The Chat Assistant solution architecture is highly flexible and customizable, but also secure: users can create private data feeds that enable the upload and indexing of proprietary company files – e.g., HR manuals, technical reports, contracts and customer transaction records – without sharing that sensitive information with the public version of Open AI. And once these docs are indexed, the Chat Bot can search and answer employee queries with real knowledge of the internal information.

But that’s just the beginning: Xebia AI Chat Assistant supercharges productivity by creating user stories and transcribing videos, writing software tests and generating test data, formulating database schema and migration scripts. The app saves time & money by leveraging AI for proofreading and summarization, writing and reviewing code, translating text into multiple languages, and generating code for rendering original imagery and ER diagrams in or Mermaid.

And to return to that teaser about choosing your "role," Xebia AI Chat Assistant accelerates your AI queries and projects with eleven pre-loaded "Personas," corresponding to varied business roles. Each Persona focuses on a likely set of business needs, and offers role-appropriate AI prompts.

Personas and Prompts

As noted, Xebia AI Chat Assistant is still at “alpha” stage, and is being used (exclusively) by 200+ Xebia engineers. But we’re on an accelerated path toward making the app available to customers, and it’s a great way to kick off a discussion about your organization’s 2024 plans for AI & ML.

Finally, let’s not forget the big picture: the World Economic Forum's "The Future of Jobs Report 2020" predicts 85 million jobs globally will be replaced by AI by 2025. The same report also indicates that AI can potentially generate 97 million new roles. At Xebia, we’re harnessing all our creative talents to help our customers create better experiences for hardworking employees. We’re using "AI on AWS" today, to advance business enterprises, build sensible data flows, and run cloud environments efficiently and in a green-friendly manner.

Stay tuned as we continue to cover AWS re:Invent 2023, and offer ‘our take’ on AI and your future in the cloud.