Built on Bedrock Demo Night
We hosted a Built on Bedrock demo night on March 18th at the AWS Gen AI loft in San Francisco.
It was a great night to see the latest in Generative AI from Bay Area startups.
The presentations included applications of Gen AI for code generation, code deployment, pull request and issue tracking, context management, and agentic AI.
Each of the startups who presented leverages Amazon Bedrock — Amazon’s fully managed generative AI service.
The evening started with a fun gaming demo that leverages Bedrock and Amazon Nova models, presented by Pat Santora and Frédéric Nowak.
Learn more about the startups and watch the demo video below.
Presenters
Codeium
Codeium Windsurf, presented by Jeff Wang, is a Gen AI IDE incorporating both coding co-pilot and agentic functionality. They have a feature, Riptide, that goes across the entire code base using LLMs and explains and summarizes what the repo does. Codeium also features the ability to run commands. For example, developers can run a command to write and run unit tests — and automatically fix the failures. Developers can easily create web apps from scratch via a simple, natural-language prompt.
Defang
Defang, presented by Lio Lunesu, is a Gen AI deployment tool for containerized applications. Developers can easily deploy via Defang, without needing DevOps intervention. Defang deploys directly from a developer’s machine to their own cloud account. It writes all the infrastructure-as-code for the developer. Defang follows best practices including the Amazon Well-Architected Framework.
Graph
Graph, by LaunchNotes, presented by Greg Dame, is a Slack-native, Gen AI agent that takes code repository tools and issue trackers, and provides a contextual layer on what the team is working on and the status. Graph supports Github, Jira, and Linear. Users can easily chat with Graph in Slack and find out things like incident analysis, pull request details, or who is the subject matter expert on a particular part of the code base.
Letta
Letta, presented by Sarah Wooders, is an agentic framework, founded by the team behind MemGPT, focused on state, memory, and context management. It works across model providers — state and context are represented in an agnostic way. Letta features an agent development environment that can be used via API or the UI. These gents can then use tools to manage their own memory, and context window. You can also create your own personas for the agents.
NinjaTech
NinjaTech, presented by Babak Pahlavan, is an agentic assistant. NinjaTech operates at the skill level and can be used for eight skills including writing, brainstorming, code generation, and image creation.
Demo Videos
Next Event
Our next event is April 16th at the Gen AI Loft in SF. Register to attend.
If you are interested in presenting, fill out the demo application.
Arte Merritt is a GTM specialist for Amazon Bedrock for Startups. He is the founder of Reconify, an analytics and optimization platform for Generative AI. Previously, he led the Global Conversational AI partner initiative at AWS. He was the founder and CEO of the leading analytics platform for conversational AI, leading the company to 20,000 customers, 90B messages processed, and multiple acquisition offers. He is a frequent author and speaker on Gen AI, Conversational AI, and data insights. Arte is an MIT alum.
