along withAWS re:Invent 2025 Conference EventsAWS has announced numerous technology collaborations, covering AI applications in various vertical sectors such as video understanding, financial payments, transportation, and smart buildings.
Among these developments, TwelveLabs launched Marengo 3.0, touted as the "world's most powerful" video understanding model, on the Amazon Bedrock platform. Additionally, Visa announced a partnership with AWS to promote "proxy payments," enabling AI services to securely conduct various transactions on behalf of users.
TwelveLabs Marengo 3.0: "Understanding" Videos Like a Human
TwelveLabs' Marengo 3.0 model, launched on the Amazon Bedrock platform, will address the pain point faced by enterprises: "unusable unstructured video data".
Unlike traditional models that only perform frame-by-frame analysis, Marengo 3.0 treats video as a complete dynamic system, directly linking it to dialogue, gestures, actions, and emotional content. The official claim is that this model can reduce storage costs by 50%, improve indexing performance by 2 times, and support up to 4 hours of video and 36 languages, allowing businesses to easily search and understand video content by searching text.
Visa partners with AWS to launch "proxy payments," using AI to help buy sports game tickets.
In the financial sector, Visa announced a partnership with AWS to combine Visa's global payments infrastructure with AWS's AI technology to enable secure "Agentic Payments".
Both sides willAmazon Bedrock AgentCoreThe platform provides open blueprints to help developers build intelligent workflows. Rubail Birwadker, Senior Vice President of Global Growth at Visa, stated that this partnership will become a layer of trust in the agent economy.
For example, users could instruct an AI agent to "buy basketball tickets for me if they're under $150," and the AI agent would then automatically compare prices, place the order, and process the payment. Partners include services such as Expedia and Intuit.
S&P Global adopts the MCP standard to enable AI agents to "understand" financial data.
On the software and data front, AWS announced a partnership with S&P Global to integrate its trusted market, financial, and energy intelligence directly into customers' AI workflows.
The key technology behind this collaboration lies in the Model Context Protocol (MCP). Through the integration of two new MCP servers, S&P Global's data can be seamlessly integrated into Amazon Quick Suite, meaning that in the future, the agent-based AI services built by enterprise users will not only be able to chat, but also handle complex financial issues.
For example, through S&P Global MCP for Kensho (an API that supports LLM), the AI agent can query Capital IQ’s financial data or transcripts of earnings calls, or obtain real-time insights into the energy market through the S&P Global AI Ready Data MCP Server.
S&P Global Chief AI Officer Bhavesh Dayalji emphasized that this will ensure that customers can access their trusted data at any time, whether in cloud platforms, LLMs, or AI agents, thereby optimizing decision-making processes.
Lyft, Nissan, and Trane Technologies are incorporating AI applications.
Other examples of collaboration include:
• Lyft:By leveraging the Anthropic Claude model and Amazon Bedrock to create an "Intent Agent," when a driver asks "My earnings aren't showing up," the AI service can directly access backend data to determine if the driver has just completed three trips and assist in resolving the issue. This has reduced Lyft customer service's average resolution time by 87%.
• Nissan:Nissan launched the AWS-based Nissan Scalable Open Software Platform, which speeds up testing of software-defined vehicles (SDVs) by 75% and supports collaboration among 5000 developers worldwide. Nissan also announced plans to integrate the more advanced ProPILOT AI system in 2027.
• Trane Technologies:Through its BrainBox AI system, the company successfully reduced energy consumption by nearly 15% in three Amazon Grocery pilot programs in North America by optimizing HVAC systems with AI. This project is expected to be expanded to more Amazon Grocery stores in 2026.
• BlackRock:Aladdin, its investment management platform, will be available on AWS, providing customers with more flexible investment decisions. It is expected to launch in the US in the second half of 2026.
