Ep048: Enhance Your Application with Generative AI

07/30/24 • 27 min

Federico Torreti and Tom Sly of AWS discuss how application developers can leverage Generative AI to enhance cross-app experiences, minimize custom integrations and increase the value-add of enhanced communication features.

Topics Include:

Federico Torreti intro: What is AWS’ vision for applications
Focus of Machine Learning & AI
Ability to build rapid prototypes, experimentation
Cambrian explosion of SaaS applications
The challenge of toggling of multiple applications
Bringing the world of disconnected apps
Data and focusing on the primary value for customer
Generating more content will raise number of alerts and potential of burnout
Tom Sly intro: Being delighted by businesses
Customers expect personalization from businesses
More than 150,000 customers leverage AWS for communications strategy
Leveraging messaging to increase customer delight
2 examples: restaurant and medical need
Session wrap up

Participants:

Federico Torreti – Head of Product – Amazon Q, Amazon Web Services
Tom Sly - Director, Special Projects, Amazon Web Services

Topics Include:

Federico Torreti intro: What is AWS’ vision for applications
Focus of Machine Learning & AI
Ability to build rapid prototypes, experimentation
Cambrian explosion of SaaS applications
The challenge of toggling of multiple applications
Bringing the world of disconnected apps
Data and focusing on the primary value for customer
Generating more content will raise number of alerts and potential of burnout
Tom Sly intro: Being delighted by businesses
Customers expect personalization from businesses
More than 150,000 customers leverage AWS for communications strategy
Leveraging messaging to increase customer delight
2 examples: restaurant and medical need
Session wrap up

Participants:

Federico Torreti – Head of Product – Amazon Q, Amazon Web Services
Tom Sly - Director, Special Projects, Amazon Web Services

Previous Episode

Ep047: Pricing Strategies for Generative AI with Sara Yamase of Simon-Kucher & Partners

Sara Yamase, Partner at Simon-Kucher & Partners, Head of Software, Internet and Media Practice shares guidance and industry examples on generative AI pricing and GTM strategies.

Topics Include:

What percentage of companies are developing Gen AI features for customers
Projected returns on Gen AI investment
Articulating the value is 1st step in building pricing and monetization strategy
Almost half of VC funding in last 5 quarters were for AI solutions
Getting through the hype cycle and trough of disillusionment
Education, Marketing, Food industry examples of companies clarifying value of generative AI to customers
Having strong data sets for identifying trends
Pricing tenets for generative AI
Indirect monetization – more usage, faster, greater productivity, etc
Standalone Generative AI software with bespoke pricing
Ability to scale pricing per customer usage
Subscription selling for generative AI software
Choosing pricing model for generative AI solutions
Summaries and session wrap up

Next Episode

Ep049: Enriching the Digital Experience with Amazon Bedrock

Today, Dr. Ratinder Paul Singh Ahuja of Pure Storage takes us through how his organization is leveraging Amazon Bedrock to enrich the digital experience for end users and increase returns for the company.

Topics Include:

Introductions
Generative AI activity within Pure Storage
Ownership and expertise of team, role of generative AI
Rapid changes require dedicated team reporting to office of CTO
Where Pure Storage is investing – Productivity gains, Customer Experience, R&D
Vision starts with the needs of the business users
GenAI Ops – Content, Voctor databases for ingestion, query strategies & security
Example – Purchase Acquisition Process
Features for a mobile experience
Configurations and logs are getting more enhanced with generative AI
Defining schema and improving large language model for more accurate responses
Security models and deployment
Summary and session wrap up
Q&A 1) How did the team get trained to use the generative AI tools and systems?
Q&A 2) How did the team reduce the hallucinations?
Q&A 3) How did you decide on which foundation models to use?
Q&A 4) The ROI of this particular use case
Q&A 5) Models working with poly-morphic evolving inputs
Session conclusion