Skip to content
  • There are no suggestions because the search field is empty.

Understanding What MayvnAI Can Access

MayvnAI is designed to help you analyse your manufacturing performance by conversing naturally with your OFS data. This article explains what data Mayvn can work with today and what falls outside its current scope.

What Mayvn can do today

Mayvn connects to your OFS line data and draws on the same information that powers the OFS Analytics dashboards. When you ask Mayvn a question, it retrieves the relevant data and uses it to generate answers, summaries, and charts.

Specifically, Mayvn can query and analyse:

  • Production performance: OEE, Availability, Performance Efficiency, and Quality metrics, broken down by time period, product (SKU), or crew.

  • Downtime and setup: Planned and unplanned downtime events, short stops, slow running, and setup time, including full reason code breakdowns across up to five levels of reason hierarchy.

  • Operator comments: Notes left by operators during production, including who wrote them, when, and the context they were attached to (SKU, crew, reason code).

  • Speed and throughput: Production speed, run speed, units in/out, and rated speed, available when data is grouped by product or crew.

  • OFS Flow form data: Submissions from your digital forms, including all captured field values and workstate information.


Mayvn can look across multiple lines in the Analytics view, or focus on a single line in the Console view. In both cases, you ask about a time window (a shift, a day, a week, or a custom period) and Mayvn works within that range.

What Mayvn cannot access today

There are categories of data that are not yet part of Mayvn's data scope. We want to be transparent about what those are and why they aren't straightforward to add.

Job-level metadata

Job specification fields, the parameters attached to individual orders or jobs as they run through your lines, are not currently available to Mayvn. We know this matters to our customers, and this section explains what's involved in getting there.

The core constraint is how AI models handle information. Mayvn retrieves data from your OFS system and passes it to the AI model, which has a finite working memory for each conversation turn. Every data point returned consumes a portion of that capacity.

Manufacturing KPIs like OEE and downtime are naturally aggregatable. Mayvn can ask for a day's performance across a line and receive a compact, information-dense response. Job metadata is fundamentally different. These fields exist at the individual job level, meaning a single day of production could yield hundreds or thousands of distinct values. Presenting all of them to the AI in a single conversation turn is not practical, and simply retrieving everything is not a scalable approach.

There is a deeper challenge underneath that. Some job metadata fields are highly summarisable in practice: a field with only a handful of distinct values across thousands of jobs can be excellent for grouping and comparison. Other fields are effectively unique per job and do not lend themselves to aggregation. Before Mayvn can make good use of job metadata, it needs a way to understand the nature of each field and decide how best to query and present it. This is a problem that requires careful engineering, and the product team is working on it.

Individual job records

Mayvn works with aggregated data rather than individual job-level records. You can ask "what was my OEE last Tuesday?" but not "show me every job that ran on Line 3 between 2pm and 4pm." Returning individual records for a meaningful time range produces result sets that exceed what the AI can process in a single conversation, and we would rather be upfront about that than deliver unreliable answers.

Real-time data

Mayvn queries historical data for defined time ranges. It does not have access to live machine state or real-time event streams. For real-time visibility, the OFS Console and dashboards remain the right tools.

Cross-installation queries

If your organisation runs multiple OFS installations across different sites or regions, Mayvn currently connects to one installation at a time. It cannot aggregate or compare data across separate OFS environments.

Configuration and master data

Line configurations, SKU master data, scheduling information, and similar reference data are not currently exposed to Mayvn.

Getting the most out of Mayvn today

Be specific about time ranges. Mayvn performs best when you give it a clear time window. "What was my OEE last week?" will produce a faster, more focused answer than an open-ended question about general trends.

Use product and crew breakdowns. Mayvn can group metrics by SKU or crew, which can be a powerful way to tie performance back to your production mix. "Which SKU had the most downtime last month?" is one of the most effective questions you can ask.

Leverage reason codes. The five-level reason hierarchy gives Mayvn deep visibility into why downtime and setup events occurred. "What were the top reasons for unplanned downtime this week?" is one of Mayvn's strongest capabilities.

Try OFS Flow data. If your site uses OFS Flow forms for quality checks, inspections, or operational logging, Mayvn can query that submission data. This is often an underused capability that can surface operational insights beyond what the standard performance metrics show.


If you have questions about Mayvn's capabilities or would like to discuss your specific use case, please contact your OFS account team.