Salesforce and AWS Expand AI Integrations Across Slack and Amazon Quick

Salesforce and Amazon Web Services (AWS) announced an expanded collaboration on September 15, 2026, introducing integrations that put Salesforce business context inside Amazon Quick, bring AWS frontier agents into Slack, expand zero-copy data access across the two platforms, and add real-time voice interoperability between Agentforce Voice and Amazon Connect Customer.

The companies said the integrations respond to three dynamics shaping enterprise AI adoption: demand for AI that can act on CRM intelligence wherever employees work rather than only inside the CRM, the challenge of connecting siloed data across cloud environments without requiring migration, and the opportunity to extend AI agents into real-time voice and team collaboration. In the scenarios the announcement describes, a sales team could prepare for a meeting in Amazon Quick with live CRM data, a service agent could route a customer issue across Agentforce and Connect Customer, and an operations team could query data across both environments.

Nick Johnston, Salesforce’s executive vice president of global partnerships, said trusted data now reaches into the surfaces teams rely on daily, with governance built in at every step. Rahul Pathak, vice president of data and AI GTM at AWS, said the integrations give customers “data access without migration, model choice at the right economics, and AI agents where their people already work.”

AI Agents Inside Daily Work Tools

Salesforce data, context, and skills are now natively available through Amazon Quick with no custom integration required. Enterprises can access pipeline health, account summaries, service cases, and operational analytics from inside Quick. The interoperability is built on the Model Context Protocol open standard and Salesforce’s headless architecture, which the companies said allows the two environments to interoperate natively.

AWS frontier agents can now work directly inside Slack, starting with AWS DevOps Agent and expanding to AWS Security Agent, AWS FinOps Agent, and AWS Partner Central agents. The agents carry full conversational context inside a channel, so a team can investigate an incident or check a security finding without switching to a separate console, according to the announcement.

Agentforce customers can now access Amazon Bedrock’s selection of foundation models, including models from Anthropic and NVIDIA, inside Salesforce agents and agentic workflows. OpenAI models, already generally available on Bedrock, are coming soon to Agentforce. The announcement describes Bedrock as offering the industry’s broadest selection of frontier foundation models and states that it enforces Zero Data Retention, so no customer data is stored or used for model training by any provider. Salesforce’s Trust Layer adds dynamic grounding and toxicity detection, and the companies said the combination brings HIPAA, PCI, SOC 2, and ISO 42001 compliance with full audit trails for regulated industries.

Russell Jukes, chief digital and information officer at DXC Technology, said the firm’s consultants and delivery teams have deployed more than a dozen AI agents and that Amazon Quick’s agentic capabilities have become integral to how those teams operate. Joshua Stern, director of GTM systems at Engine, said access to Bedrock’s foundation-model selection through Agentforce gives his teams the flexibility to choose the right model for each use case. Muddassar Shaikh, senior vice president of engineering at GoodRx, said bringing the AWS DevOps Agent into Slack conversations lets his teams “spot and resolve issues in their development environments faster, without breaking their flow.” Giriraj Jain, vice president of engineering services at Quicken Inc, said the DevOps Agent and Slack integration creates a shared workspace where engineers can collectively direct, refine, and act on investigation results.

Zero-Copy Data Access and Informatica MCP Servers

Salesforce Data 360 zero copy now extends to AWS Glue-managed Apache Iceberg tables, S3-backed Apache Iceberg tables, Amazon Aurora, Amazon RDS, and SageMaker Lakehouse. The expansion builds on existing support for Amazon Redshift and Amazon S3 with private connectivity, and the companies said joint customers can ground AI agents in enterprise data without moving or duplicating it and without changing the security and governance controls established in either environment. Salesforce and AWS are also exploring bidirectional semantic-information sharing with the AWS Glue Data Catalog.

Informatica is expanding general availability of its MCP servers with headless capabilities covering platform administration, catalog discovery and enrichment, data-quality scores, and master-data retrieval. The integrations are accessible from Amazon Bedrock AgentCore and Amazon Quick, which the announcement said lets enterprises deploy AI agents grounded in data that is clean, discoverable, trusted, and governed.

Voice Interoperability and Slack Dictation on AWS Trainium

With Agent2Agent support, Agentforce Voice and Amazon Connect Customer will communicate directly through real-time, bidirectional audio over WebSockets, an approach the companies describe as a new method for secure communication between AI agents across voice platforms. The companies said the support gives customers greater choice and enables AI agents to coordinate more seamlessly across the two solutions.

Separately, Salesforce AI Research worked with AWS to optimize Salesforce’s proprietary speech models on AWS Trainium chips, enabling voice dictation in Slack. Customers can capture and share ideas by voice with near-instant transcription without leaving a Slack thread, and Salesforce said running the models on purpose-built AWS silicon lowers inference cost, with the benefit passed on to customers.

Availability and Procurement

Salesforce Clouds are available in AWS Marketplace across 33 countries, and customers can apply eligible AWS spend commitments toward Salesforce products, including Agentforce, an arrangement the companies said consolidates procurement and billing. The announcement said the new integrations build on a decade-long collaboration between Salesforce and AWS.

Available now are Amazon Quick with Salesforce’s headless architecture, AWS DevOps Agent in Slack, Agentforce model choice through Amazon Bedrock, the Data 360 zero-copy expansion, the Informatica headless expansion, and Slack voice dictation built with AWS Trainium. The remaining AWS frontier agents in Slack — AWS Security Agent, AWS FinOps Agent, and AWS Partner Central agents — along with the Agentforce Voice and Amazon Connect Customer multi-agent orchestration with Agent2Agent support, are scheduled for fall 2026.