TLDR: Skai, an AI-driven advertising and analytics platform, has significantly enhanced its customer insights by adopting Amazon Bedrock Agents to power its new generative AI assistant, Celeste. This innovation allows users to analyze complex advertising data in natural language, reducing report generation time by up to 75% and providing actionable recommendations in minutes, addressing previous challenges of manual reporting, data summarization, and recommendation generation. The partnership with AWS and the use of Bedrock’s managed services enabled rapid development, ensured data privacy, and delivered substantial business improvements and customer satisfaction.
New York, NY – September 8, 2025 – Skai, a leading AI-driven omnichannel advertising and analytics platform, has announced a groundbreaking transformation in its customer insights capabilities through the strategic adoption of Amazon Bedrock Agents. This collaboration with Amazon Web Services (AWS) has led to the development of “Celeste,” a generative AI agent designed to revolutionize how brands and agencies access, analyze, and act upon their vast advertising data.
Before this innovation, Skai’s customers, primarily campaign manager teams, faced significant hurdles in extracting meaningful insights from their data. They spent days, sometimes weeks, manually preparing static reports from tables and charts, struggling to query complex datasets, and lacking intuitive visualization tools. This often left critical business questions unanswered and valuable data potential untapped. The company identified three core challenges: time-consuming report generation, difficulty in summarizing complex datasets, and the need for more data-driven recommendations.
To overcome these obstacles, Skai partnered with AWS to build Celeste, an AI assistant capable of autonomously performing complex, multi-step tasks by orchestrating calls to APIs. “Through the partnership with AWS and adoption of Amazon Bedrock Agents AI assistants that can autonomously perform complex, multi-step tasks by orchestrating calls to APIs, we’ve redefined what’s possible,” stated the authors of the AWS blog post, including Lior Heber and Yarden Ron of Skai.
Celeste empowers users to interact with their advertising data using natural language queries, eliminating the need for specialized database expertise. For instance, a user can ask, “Compare ad group performance across low-performing campaigns in Q1,” and Celeste will automatically gather and join data from various sources like profiles, campaigns, ads, products, keywords, and search terms across over 100 publishers and retail networks. Beyond simple data retrieval, Celeste synthesizes this information into cohesive narratives, identifies key patterns, and provides actionable recommendations, complete with detailed analytical approaches and ready-to-present stakeholder materials.
A compelling example highlighted in the report involved a query about launching a new home security product. Celeste quickly identified high-performing keywords and their match types from existing campaigns, provided estimated CPCs, budgets, and a high-level testing plan, including negative keywords to avoid conflicts. This demonstrates Celeste’s ability to perform exploratory analysis, summarize trends, and offer intelligent recommendations.
A critical factor in Skai’s success was the choice of Amazon Bedrock as the foundational AI service. Unlike alternatives requiring extensive infrastructure setup, Bedrock offered a “frictionless path from concept to production.” Skai could rapidly experiment with multiple foundation models (FMs) like Anthropic’s Claude, Meta’s Llama, and Amazon Nova, iterating quickly through Amazon Bedrock Agents without managing separate environments. “We went from whiteboard to a working prototype in a single sprint. With traditional approaches, we’d still be configuring infrastructure,” noted one developer. This managed service minimized DevOps overhead, allowing Skai to focus on customer value and data integration while ensuring client data remained private and secure.
AWS Enterprise Support played a crucial role, acting as a strategic partner. Dedicated Technical Account Management (TAM) and Solutions Architect (SA) services provided proactive guidance, architectural reviews, and direct access to AWS service teams, which was vital for navigating complex generative AI challenges. This comprehensive support helped Skai reduce its proof-of-concept to production timeline by 50% and maintain 99.9% uptime during critical demonstrations.
The solution architecture involves a custom layer (Customer Experience UI, Chat Manager, Chat Executor) interfacing with the Amazon Bedrock agent, which utilizes Anthropic’s Claude 3.5 Sonnet V2 as the foundation model. A custom Tool API queries isolated customer data, ensuring privacy and compliance. Skai confidently assures clients that their data is not used to train any models, a key differentiator in the market.
Skai also successfully overcame critical challenges, including initial latency issues of 90 seconds or more, which were reduced by 30% through a custom orchestrator and streaming. Dynamic session chunking and Retrieval Augmented Generation (RAG)-based memory retrieval addressed limitations with large token windows for multi-year datasets. Robust error handling, including real-time tracing with Amazon CloudWatch Logs Insights and automatic retry mechanisms, ensured reliability at scale.
The business results are compelling: Report Generation Time: 50% Faster Case Study Generation Time: 75% Faster QBR Composition Time: 80% Faster Report to Recommendation Time: 90% Faster
Customer testimonials underscore the impact. Aram Howard, Amazon Advertising Executive at Channel Bakers, stated, “It’s made my life easier. It’s made my team’s life easier. It’s made my clients’ lives easier and better… being able to do that as efficiently as possible and as fluidly as possible with Celeste AI is always a welcome addition to Skai.” Erick Rudloph, Director of Search Marketing at Xcite Media Group, added, “Celeste is saving hours of time. It’s like having another set of eyes to give suggestions. I’m so stoked to see where this could take us.”
Looking ahead, Skai plans to expand Celeste’s capabilities by personalizing user experiences, allowing ingestion of custom data assets, and integrating new tools for seamless team collaboration, including generating client presentations and data dashboards.
Also Read:
- AWS and Gradial Partner to Revolutionize Enterprise Content Operations with Amazon Nova AI
- AWS Unveils Integrated Solution to Accelerate AI Agent Adoption in Education with Strands Agents, Bedrock AgentCore, and LibreChat
In conclusion, Skai’s implementation of Amazon Bedrock Agents has not only transformed raw data into strategic assets but also provided a blueprint for other organizations seeking to democratize data analytics. The ease of transitioning from concept to production with a fully managed service allowed Skai to focus on its core strengths, delivering significant business acceleration and maintaining enterprise-grade security, performance, and reliability.
