Brainbox AI: Revolutionizing Facility Management with Amazon Bedrock | AWS OnAir NY Summit 2024

Brainbox AI: Revolutionizing Facility Management with Amazon Bedrock | AWS OnAir NY Summit 2024

Hey everyone, thank you so much for hanging out with us today and coming back. My name is Adrian San Miguel, I am a Partner Enterprise Architect here at AWS, and I'm joined for the very first time, finally, by my colleague Jillian. Say hi, Jillian.

Hi everyone, SL Architect, and I'm really excited here for the summit. But the real stars of today's segment are the folks from Brainbox AI. So let's do some intros.

Meet the Team

Hey everybody, my name is Har and Codé, I am the Chief of Staff at Brainbox AI.

Hello everybody, Jean-Simon Venne, one of the founders of Brainbox AI, and presently playing the role of CTO. Somebody's got to do it, right?

Introduction to Brainbox AI

For those in the audience that may not have heard of you or worked with you, Jean-Simon, in your own words, can you help us understand what Brainbox AI is all about?

Absolutely. Brainbox AI is like the autonomous vehicle technology but for buildings. We connect to buildings, extract the data, and operate them in real-time without any human intervention, aiming to improve their performance in terms of energy saving and emissions reduction. You have to remember that buildings account for about one-third of the total energy consumed on our planet. We are able to reduce that energy consumption by 20 to 25%.

Leveraging Amazon Bedrock

What's really exciting, and I'm sure the folks want to learn more about, is your use of generative AI and Amazon Bedrock. Can you explain how you're using Amazon Bedrock?

We’ve been leveraging Amazon Bedrock since last summer as Beta users. What's remarkable about Bedrock is its Carousel of models that you can interchange depending on the question you're trying to address. We connected Bedrock to our existing knowledge base at Brainbox AI and built an autonomous AI agent. This agent taps into our knowledge to significantly improve troubleshooting efforts. For example, tasks that would take a person 1-2 hours to diagnose, we can now achieve in 2 minutes using Bedrock.

Real-World Applications

As an old data center infrastructure guy myself, I used to spend a lot of time troubleshooting HVAC issues, humidity points, and more. Such a system could drastically streamline these processes. How else does your generative AI help facility managers beyond just break-fix scenarios?

Generative AI opens endless possibilities for interacting with Brainbox AI's technology. Our initial product was a predictive autonomous AI layer that optimizes building efficiency. Generative AI allows end users like facility managers or service technicians to interact with these systems to gain valuable insights. For example, a facility operator can ask Arya, our AI assistant, "What do I need to know about my building today?" Arya can generate insights, provide data in various formats, and even execute certain changes like adjusting set points or addressing HVAC error codes.

Demo Time

Our audience would love to see this in action. Can you show us a quick demo?

Absolutely. Here's a live demo of Arya.

Arya: "I'm preparing my yearly sustainability report. Can you give me an overview of my scope 1 and 2 emissions?"

Arya: "Here's the breakdown of your portfolio emissions."

User: "Can I get an Excel file with a detailed breakdown of my building’s emissions and carbon intensity based on grid emission rates?"

Arya: "Sure, here it is."

User: "What GHG emissions were mitigated thanks to Brainbox AI this year?"

Arya: "Brainbox AI reduced 2,579 tons of CO2 across your portfolio this year."

User: "Can you compile all this information into a report and email it to me?"

Arya: "Yes, the report has been sent."

Implementation of Bedrock

That was impressive. It showcases your additional machine learning capabilities well. John, can you talk about the AWS services besides Bedrock that are integral to Arya's success?

Sure. We're talking here about an autonomous AI agent where LLM is just one component. We use various AWS services like Lambda workers and Fargate to beef up the LLM. For example, Arya can be given missions to monitor specific parameters and execute commands autonomously while reporting back. We also can capture the knowledge from skilled human operators and integrate it into Arya’s knowledge base for future references.

Productivity Gains

It's fascinating to learn about the productivity gains. Facility managers often feel understaffed, so this can indeed help them accomplish more. What are the most common features facility managers use Arya for?

Typically, facility managers receive a large number of alerts daily. They prioritize and solve as many as possible but seldom complete all tasks. Arya helps them clear the entire list of issues, substantially increasing their productivity and ensuring the building operates impeccably.

Wrap-up and Future Directions

Fantastic. This has been fascinating. Where can people go to learn more about Brainbox AI?

Our website, brainboxai.com, has the most up-to-date information about everything we are doing.

Thank you so much for joining us today. Everyone, stick around—we've got more coming up at the AWS Summit, including the keynote, so stay tuned.

Keywords

• Brainbox AI
• Amazon Bedrock
• Generative AI
• Facility Management
• Energy Efficiency
• Autonomous AI Agents
• Sustainability
• HVAC Optimization
• AWS Summit

FAQ

Q: What is Brainbox AI? A: Brainbox AI is an autonomous technology for buildings, aimed at improving energy efficiency and reducing emissions by operating in real-time without human intervention.

Q: How does Brainbox AI leverage Amazon Bedrock? A: Brainbox AI uses Amazon Bedrock's Carousel of models to interchange and optimize data processing, enabling real-time troubleshooting and energy efficiency improvements.

Q: What are the main features of Arya, the AI assistant? A: Arya provides insights, generates reports, adjusts set points, addresses HVAC error codes, and can execute autonomous commands based on predefined missions.

Q: How does Brainbox AI impact productivity for facility managers? A: Arya helps facility managers clear their list of issues more efficiently, often completing tasks much faster than manual methods, thus significantly increasing productivity.

Q: Where can I learn more about Brainbox AI? A: More information can be found on their website, brainboxai.com.