Autonomous Building Technology Enhances Energy Efficiency and Sustainability with AI

Artificial Intelligence & MLOps
Data Modernization & Analytics

Solution Implemented

Contributed to the development of ARIA, an AI-powered virtual building assistant for intuitive HVAC and building data analytics and operations management.

Utilized multiple models on Amazon Bedrock, including Anthropic’s state-of-the-art model Claude for the majority of workloads, to deliver a conversational interface and generative AI capabilities.

Deployed solution on serverless AWS cloud infrastructure through infrastructure-as-code, architected for performance, scalability, and security.

Outcomes Experienced

Streamlined data analysis, decision-making, and HVAC code deployment, reducing complexity and manual effort.

Provides seamless access to BrainBoxAI’s core technology that enables building operators to reduce energy costs by up to 25% and greenhouse gas emissions by up to 40%.

Transforming Energy Management with AI and Cloud Technology

Energy utilization is a core problem when it comes to sustainability and climate change. According to the International Energy Agency (IEA), buildings account for 30 percent of global energy consumption and 26 percent of global energy-related emissions. Driving towards sustainability demands the examination of built environments, and they represent a significant opportunity for climate change mitigation. 

BrainBox AI is an energy management and technology optimization company that works with commercial real estate to improve the efficiency and sustainability of their HVAC operations. The company supports real estate clients in various sectors, including office buildings, hotels, commercial retail, grocery stores, airports, and more.

BrainBox AI identified unique opportunities across the built environment to improve energy efficiency and minimize the operating expenditure traditionally associated with such efforts. Leveraging autonomous AI, BrainBox AI makes buildings smarter, greener, and more efficient, helping reduce their impact towards energy consumption and climate change. Some of their most notable clients include Dollar Tree, Holiday Inn, GWL Realty, Sleep Country, Brisbane Airport and other large North American retailers who have begun implementing their systems across thousands of their locations.

With a portfolio of over 14,900 retail locations, BrainBox AI’s systems deal with vast amounts of data. The company approached Caylent to develop an Artificial Responsive Intelligent Assistant (ARIA), to revolutionize HVAC management.

BrainBox AI’s ARIA shows tremendous promise towards accelerating and improving the way their customers understand and make decisions around their HVAC and energy management systems. Its ability to provide complex data analytics and visualizations, optimization recommendations, and answers to facility management questions in real-time with little to no manual effort, maximize the efficiency of operations and provide a standardized set of capabilities to all end users regardless of their levels of expertise.

End-users can query via voice or text the vast data landscape and obtain insights, complete with graphs and charts instantly, on their desktop or mobile device. They can take a proactive approach in identifying and mitigating problems early while continually improving system efficiency. The solution can query BrainBox AI's data lake and answer questions that will drive well-informed technical and business decisions and also provide recommendations, helping decision makers mitigate risks, optimize cost efficiency and drive energy usage responsibly towards a more sustainable future.


The company approached Caylent to develop ARIA, a natural language interface powered by AWS’s GenAI suite that improves insight access and depth, enables users to interact conversationally for data analysis and command HVAC and other building operations systems. 

With ARIA, BrainBox AI aims to offer a seamless conversational experience to tackle even the most complex operational challenges. For usage analysis and emissions reporting, imagine asking an AI assistant to plot all of the energy consumption from every retail store on the East Coast and having that information returned to an operator instantly with the ability to seamlessly refine and finalize that data into a standard compliance reporting format - all from just asking a few questions. For building operations and energy savings, imagine being able to automatically make HVAC system changes intuitively, as heating and cooling needs shift with weather and occupancy factors.

ARIA’s design is intended to help facilities managers and building operators seamlessly command their building's operations, from RTUs to chillers, boilers, and more, through voice or text - transforming building data into precise insights and strategic actions in real-time.


Caylent’s data and AI experts worked collaboratively with BrainBox AI’s CTO, data scientists, engineers, and DevOps specialists to understand BrainBox AI’s goals and plan out the deployment of AWS tooling for data ingestion, report generation, and conversational interfacing with BrainBox AI devices, HVAC systems, and other relevant building data. 

Leveraging Caylent’s deep experience in Amazon Bedrock, industry-leading LLMs, including Anthropic’s Claude for the majority of workloads, and production-grade generative AI deployments, ARIA was built with scale, speed, and security in mind. 

ARIA leverages a serverless data lake platform built on Amazon S3 and Amazon Athena, where historical data from BrainBox AI devices is stored. It also reaches out to BrainBox AI’s real-time APIs and databases for current building readings. By combining these two data sources ARIA can intelligently query, analyze, and report on building performance.

ARIA is built on Amazon Bedrock, powered by its multi model capability, and uses these models to identify the user’s intent from their input and performs additional Bedrock LLM inferences to build a detailed answer using BrainBox AI ’s data lake and APIs. Bedrock models provide the capability to generate simple natural language summaries and rich data visualizations such as time-series graphs and charts.

ARIA was architected as infrastructure-as-code using Terraform, enabling seamless deployments for user onboarding.

Adding to the capabilities of this technology, BrainBox AI and Caylent are enabling automation and activation functionality that will allow the solution to deploy changes across HVAC systems in line with each end-user’s unique efficiency targets. ARIA can take in simple language requests, interface with HVAC tooling in the languages they operate in, and deploy code changes automatically without requiring sophisticated tooling experience from its user.

Related Services

Caylent Catalysts™

Generative AI Strategy

Accelerate your generative AI initiatives with ideation sessions for use case prioritization, foundation model selection, and an assessment of your data landscape and organizational readiness.

Caylent Catalysts™

Generative AI Knowledge Base

Enhance access to your corporate assets with a custom AI chatbot powered by Anthropic Claude 3 in Amazon Bedrock. Our AI experts will deploy a production-grade prototype to your AWS environment and configure Amazon Kendra to index your data.

Accelerate your GenAI initiatives

Leveraging our accelerators and technical experience

Browse GenAI Offerings

View our case studies

Smart Eye Technology Logo

Smart Eye Technology

Digital Privacy SaaS company reduces the Total Cost of Ownership by adopting cloud-native infrastructure

Read more
Criteria Logo


Talent success company migrates from Snowflake to AWS with Amazon Redshift Serverless

Read more
Upside Logo


Retail technology company improved retail promotion and recommendation engine with MLOps

Read more

Related Case Studies

Trulioo Logo


Identity Verification Provider Uses Generative AI to Augment Customer Onboarding Experiences

Read more
FloSports Logo


Digital Sports Streaming Platform Enhances Fan Experiences on AWS

Read more