Revenue Strategy & Data Analytics Lead

Tulip introduces new AI Authoring features to accelerate solution building. AI Authoring builds on Tulip's composable platform, enabling AI-powered planning, building, and review to accelerate solution development even further, so teams can spend more time on what matters most: continuous transformation. Solutions go live faster and stay consistent across sites, with the engineer in charge of what ships. BOSTON, Mass., October 9, 2026 - At Operations Calling 2026, Tulip, the leader in Frontline Operations, introduced AI Authoring, new capabilities that bring AI Agents into planning, building, and reviewing solutions. Projects turn a team's existing context and best practices into a plan for the whole production system. The Authoring Agent, a chat panel in the App Editor, drafts new apps and changes existing ones once the engineer approves its plan. Composable AI Agents give reviewers a first pass on testing, documentation, and evaluation. For teams that prefer to build in code, Code-Based Apps let engineers use their own IDEs and AI coding agents on the same governed platform. Projects and the Authoring Agent are entering customer testing, Composable AI Agents are in open beta, and Code mode is previewing. AI has made creating software fast, but on a factory floor, software built without the line's data, permissions, and standards lands on the engineers who have to review and fix it before it can run. AI Authoring builds on Tulip's composability and no-code tools, giving AI the context of the production system so what it builds fits the operation and stays governed as it grows. "Over the past year, AI has powered an explosion of the number of people who can write code. At Tulip, we see that as further validation of our day one vision. Tulip was founded on the belief that the people closest to operations fundamentally lacked the tools that they needed to solve their biggest challenges." said Mason Glidden, Chief Product and Engineering Officer at Tulip. Projects: plan from your existing context and best practices. Projects starts with what a team already knows. Engineers bring in the SOPs, layouts, and process documents behind the work, along with their existing apps, libraries, and best practices. From that context, Projects proposes a plan for a value stream, line, or process, so teams start from how they actually operate instead of a blank page. Engineers can also ask the project plain-language questions, such as where a given SOP step is implemented. A better plan means faster execution and better long-term outcomes. The plan lives on an operational canvas that shows how apps, automations, machines, and agents work together, so teams can see what a change will touch before they build it. A solution can then be versioned and deployed as a single unit, which makes rolling it out to the next site a deployment rather than a rebuild, and live production data returns to the project to inform the next change. Projects is now entering customer testing. The Authoring Agent: builds the way an engineer would. The Authoring Agent lives in the App Editor, as a chat panel where engineers work with it directly, on top of the data model their team has already defined. It starts where engineers already start, with the SOP. An engineer uploads one or describes the process, the agent asks the clarifying questions a good engineer would ask, and then it proposes a build plan. Nothing changes until the engineer approves it. From there the agent drafts the app's steps and logic on that existing data, rather than inventing a new data model, and reports what it changed. It can edit apps a team already runs, not only new ones, and it draws on Tulip's operations expertise as well as the team's own context, so drafts land even closer. Because the result is a standard no-code app, the engineer reviews each draft, takes over in the App Editor at any point, and decides what gets published. The Authoring Agent builds on AI Composer, which Tulip introduced in 2025 to turn work instructions into step-by-step guidance apps, and it starts with apps, with more of the platform to follow. The Authoring Agent is now entering customer testing. Code-Based Apps: Generate code, run on the same platform. Code-Based Apps let engineers build custom apps with their own IDEs and AI coding agents, then run them in Tulip. Code fits use cases that need more flexibility than no-code, such as interactive visualizations, complex logic, work instructions that vary across many product configurations, and apps that pull live data from systems like PLM and ERP. Code-Based Apps call Tulip's connectors, tables, and widgets rather than reimplementing them, so no passwords, OAuth flows, or hardcoded keys end up in app code. The platform enforces site portability, operator attribution, and zero-trust IT security, and carries Git pull requests to the shop floor as frozen, fully audited versions designed for GxP environments. Code-Based Apps are now previewing, and Tulip is accepting design partners. Composable AI Agents: review for quality and consistency. Building faster only helps if review keeps pace. With pre-built composable AI agents, teams get a first pass on testing, documentation, and evaluation, so reviewers can spend their time on judgment. Teams can tailor each agent with their own deployment context and enterprise best practices, which keeps quality consistent across everyone who builds. Composable AI Agents were announced last year at Operations Calling and are now in open beta. Governed like everything else in Tulip. AI Authoring runs inside the Tulip platform, so it inherits what IT and Quality already require: a defined data model, role-based access control, review and publishing gates, versioning, and audit trails, backed by a security posture built for regulated manufacturing. Everything the Authoring Agent builds attaches to the same data model, so more solutions, built by more people, add up to one system instead of a collection of islands. Composable AI Agents carry a team's standards into review, so consistency comes from the system rather than from a stretched team checking every solution by hand. Projects, the Authoring Agent, and Code-based mode are just a few of the product announcements Tulip made at Operations Calling 2026. Manufacturers can apply for early access through the Tulip AI early adopter program. To learn more, read the full Operations Calling 2026 announcement. About Tulip. Tulip is the Frontline Operations Platform. Manufacturers in regulated and complex industries use Tulip to build apps, agents, and automations in no-code, with AI, or in code, and to run them on one governed, composable platform with end-to-end traceability and the full context of the operation. A spinoff from MIT, Tulip is headquartered in Somerville, MA, with offices in Germany, Hungary, Singapore, Israel, and Japan.

Tulip connects machine data, AI Agents, and factory video through a common operational record. October 9, 2026 By Craig Resnick Company and Product News Tulip has introduced operational intelligence capabilities that connect machine data, AI agents, enterprise systems, and factory video through a common operational record. Tulip has announced a set of operational intelligence capabilities intended to help manufacturers run production and orchestrate people, machines, and AI agents from a common operational record. The additions include Industrial Connectivity, Native MCP Connectors, Composable AI Agents, and expanded Factory Playback capabilities. The announcements address a persistent challenge in manufacturing operations. Machine data, work instructions, order information, operator actions, video, and enterprise-system records are often distributed across separate systems. As a result, investigating a quality issue or production interruption can require teams to assemble evidence manually before they can determine what happened and what action should follow. Connecting machine data to operational context. Tulip's Industrial Connectivity offering is designed to bring machine data from existing drivers, historians, and unified namespace environments into the platform. The company said the capability can normalize incoming data, convert units, derive machine states from signals, and map equipment to common data types. The more significant element is the operational context added around that machine information. Rather than treating machine signals as isolated time-series data, Tulip connects them to the associated order, process step, station, and operator. This can help teams understand not only what a machine reported, but what work was taking place, what decision was made, and what occurred next. For manufacturers with equipment from multiple vendors and generations, this type of contextualization is essential. Bringing data into a common namespace alone does not guarantee that the information is usable for analysis, workflow execution, or AI-assisted decision-making. Extending the record to enterprise systems and AI tools. Tulip also previewed Native MCP Connectors, which use Model Context Protocol (MCP) to enable two-way data exchange between the Tulip platform, enterprise systems, and AI tools that support the protocol. Under this approach, Tulip agents can retrieve and use information from connected systems, while external AI assistants and coding tools can access Tulip data through its MCP server. The company said these interactions operate under the permissions established in Tulip, allowing organizations to control whether a connected tool can view information or make changes to the record. This is particularly relevant as manufacturers move beyond stand-alone AI pilots. An AI agent can only support a production workflow effectively when it has access to the right operational context and operates within defined permissions, approval requirements, and audit trails. Seeing what happened on the production line. Tulip has also expanded Factory Playback, which places video from existing factory cameras on the same timeline as machine and order data. Engineers can examine what was happening on the line at the time of an event, rather than relying solely on system records or later recollections. New human activity detection capabilities are intended to support process analysis by identifying how time is spent at a workstation, whether a procedure was followed, and whether an assembly activity occurred as expected. Tulip said Factory Playback can also initiate work based on detected events, creating a link between observation, analysis, workflow execution, and documentation. Together, these capabilities are intended to reduce the need for teams to piece together evidence from separate systems during root-cause analysis. They also provide a more complete record for testing changes, confirming outcomes, and tracing decisions made by people or AI agents back to the relevant machine data, video, and order information. Related ARC insights. Tulip's operational intelligence updates reflect the broader industrial requirement for AI systems to be grounded in contextualized operational data. As manufacturers connect AI agents to frontline workflows, the ability to combine data, execution context, permissions, and traceability will be as important as the AI model itself. Read ARC's perspectives on: For manufacturers, the potential value lies in creating a shared operational record that supports both real-time execution and retrospective analysis. This can give frontline teams and AI-enabled systems a common basis for understanding production events, acting within defined controls, and improving operations over time.