Sema4.ai’s autonomous agent building platform gets easier to use, adds deeper business context and more

Sema4.ai Inc., a startup that provides tools for building and managing artificial intelligence agents, today announced a major overhaul of its platform, with major changes to every layer of the agent development stack.

The company says the revamped platform improves everything from how agents are built, to how they see and understand business context, to how they are deployed within a customer’s computing environment.

Sema4 started gaining attention last year when it raised $25 million in a Series A round. It was founded by CEO Rob Bearden, who previously held the same role at big data company Cloudera Inc.

We provide a platform that allows non-technical employees to create AI agents with natural language prompts. Users can build the agents they need and configure them to perform tasks such as extracting information from documents. Agents can also use external software applications through a set of integrations called “actions.”

Users create AI agents using an interface called Work Room. This gives your agents access to all the AI ​​capabilities they need. However, a second tool called Control Room allows administrators to monitor these agents.

Sema4 says today’s update seeks to fix many of the issues that have caused agent AI deployments to stall instead of scale, including fragmented systems, disconnected data, and tools that make more sense to software developers than Sema4’s actual target users.

For example, Sema4 has introduced a new Agent Builder tool that requires no technical expertise. It has a more organizational context, allowing users to easily explain by voice or input, or upload standard operating procedures that they would like to automate. Agent Builder generates functional agent runbooks that enable autonomous agents to perform their specific tasks.

In addition to support for voice, text, and document input, users will also have access to a broader library of pre-built agent skills, allowing agents to acquire “persistent memory.” This means you can learn from experience, retain fixes, surface workflow recommendations, and essentially enhance your organization’s knowledge over time.

There’s also a new MCP Access Gallery, which enables Sema4 agents to leverage third-party software such as Snowflake, Slack, GitHub, Google Workspace, and HubSpot through the Model Context Protocol. Finally, Agent Builder gains support for federated and validated queries, allowing users to ask one question and search for the answer across all databases, spreadsheets, and enterprise systems that the AI ​​agent has access to.

Sema4 also reimagined agent context with a new business context layer to help you quickly understand how your enterprise data is connected across multiple databases, systems, and workflows. One of the key features is a business ontology that allows you to map relationships between individual customers, purchase orders, invoices, shipments, etc. This allows agents to reason across the entire business, rather than being limited to a single database or system.

Meanwhile, there are new semantic layer enhancements within the business context layer that help improve performance for agents working across data silos.

“This release dramatically makes it easier to build and deploy enterprise AI agents while providing a deeper understanding of how businesses actually operate,” said Paul Codding, co-founder and senior vice president of product and customer experience. “Agents need to be able to reason and adapt to business concepts, not just columns and rows.”

Architecturally, this release introduces improvements that extend platform availability and simplify agent deployment across Amazon Web Services, Google Cloud Platform, Microsoft Azure, and Snowflake.

“We are making enterprise AI agents more accessible to business users who understand their operations, more connected to the systems where the data resides, and more reliable for enterprise operations,” Codding added. “All the improvements compound: agents that are easier to build deploy faster, agents with deeper business context provide more accurate results, and agents that run on simpler architectures reach production faster.”

Image: Sema4