Product, People, Partnership: What Microsoft’s Frontier Vision Looks Like on the Plant Floor from a Partner’s Perspective

Two pieces of news landed within weeks of each other this summer. This July, Microsoft named Vishvesh Shah and Parth Desai Most Valuable Professionals for the second year running, placing them among just five Microsoft MVPs in the entire United States recognized in the IoT category. Then, on August 4, Microsoft certified Litmus Edge Manager’s Site Model under the Microsoft AI Cloud Partner Program (MAICPP) Solutions Partner with certified software designation for Manufacturing AI. It would be easy to read these as two unrelated announcements: one about people, one about a product. At Litmus, we see them as the same story, told twice.

Both also arrived in the same timeframe that Microsoft laid out where it believes enterprise AI is headed. On July 2, Microsoft launched Microsoft Frontier Company, a $2.5 billion commitment that mobilizes more than 6,000 industry, engineering, and AI experts to work inside customer organizations, designing, building, deploying, and operating AI where the work actually happens. Microsoft’s framing is AI engineering that “amplifies and protects your intelligence”: AI that makes a customer’s own expertise more valuable, while keeping that expertise, data, and IP firmly in the customer’s hands, with intelligence and trust. Reading that announcement, we recognized a great deal of our own playbook, and it’s the lens we’ll use to tell this story.

The Manufacturing AI certification didn’t appear out of nowhere. Litmus and Microsoft have spent the last few years deepening a technical relationship. In March 2025, at Hannover Messe, the two companies announced a strategic partnership centered on helping manufacturers address a key challenge to unify their edge-to-cloud data management, securely, delivering AI-ready data. By integrating Litmus Edge with Azure IoT Operations through the Akri Litmus Connector, manufacturers gain unified edge-to-cloud data management across Azure Arc, Azure Resource Manager, and Microsoft Entra ID. Litmus CEO Vatsal Shah framed it as a way to help industrial companies simplify the complexity of edge data management as they pursue digital transformation, while Microsoft’s Dayan Rodriguez, Corporate Vice President for Manufacturing and Mobility, pointed to Azure’s adaptive cloud approach unifying information across edge, on-premises, and cloud environments.

That integration kept expanding. By November 2025, Litmus and Microsoft had extended the Akri Litmus Connector. The expansion helps manufacturers accelerate and automate industrial data onboarding and modeling directly inside Azure IoT Operations, providing real-time device and asset discoverability, seamless onboarding, and Azure-native representation of Litmus Edge data. Litmus is positioned as an independent software vendor inside Microsoft Cloud for Manufacturing and Litmus Edge is also available, and IP co-sell eligible, on Azure Marketplace.

The MAICPP Solutions Partner for Manufacturing AI certification is the natural next rung on that ladder. To earn it, Litmus Edge Manager had to pass Microsoft’s technical audit demonstrating real AI capability on Azure, meet customer-success thresholds over a trailing twelve months, and submit referenceable case-study evidence specific to manufacturing. It’s Microsoft’s formal acknowledgment that the software itself, not just the relationship, clears the technical bar for Manufacturing AI.

It’s also worth noticing what that bar measures. Customer-success thresholds and referenceable case studies are outcome tests, not experiment tests. That tracks with the shift Microsoft describes at the heart of Frontier Company: customers have moved past AI experimentation and now expect measurable business results. In manufacturing, that means AI that survives contact with a real production line.

Here’s the kind of problem that certification is actually about. Picture a manufacturer running a dozen plants, each with some version of the same CNC line or packaging machine, installed over fifteen years by different integrators. On paper, it’s the same equipment doing the same job. In the historian at each site, it looks nothing alike: one plant tags a spindle fault as “Alarm_12,” another calls the identical fault “E-Stop-Spindle,” and a third never logs it at all. Multiply that by every sensor, PLC, and protocol on the floor, and a plant manager trying to compare downtime across sites, or feed that data into an Azure AI model to predict the next failure, is stuck translating by hand before any analysis can start.

That’s a data-modeling problem before it’s an AI problem, and it’s exactly the unglamorous, site-specific work that decides whether an edge-to-cloud platform like Azure IoT Operations ever gets past a pilot. Solving it usually means someone technical actually walking the floor, plant by plant, mapping out where the naming diverges and where the protocols don’t line up. It’s hands-on engineering work that never shows up in a product demo, but it’s exactly what separates a working deployment from a stalled one.

Microsoft makes the same argument in its Frontier announcement: enterprise AI that keeps improving requires deep industry knowledge, not just general-purpose engineering talent. On the plant floor, industry knowledge looks like knowing that “Alarm_12” and “E-Stop-Spindle” are the same event, and knowing why it matters that the third plant never logged it.

That’s the work behind Vishvesh and Parth’s MVP recognition. It wasn’t lab work. It was two years of close, in-the-field collaboration with Microsoft developing a Litmus Edge Bridge for Azure and Azure IoT Operations: showing up at a customer’s site with exactly this kind of tangled, plant-specific naming problem, building the connector and data model that got Azure IoT Operations and Litmus Edge to agree on what “Alarm_12” actually meant, and then carrying what worked back into the product and the Microsoft community through knowledge-sharing, product feedback, and real-world implementation.

Microsoft built Frontier Company on the premise that the fastest path from AI ambition to AI results is putting engineers inside the customer, side by side with the people who run the business. Microsoft’s Judson Althoff described the model as going beyond what the industry has labeled Forward Deployed Engineering. Vishvesh and Parth have been living a manufacturing version of that idea for two years: embedded at the plant, accountable for a working outcome, and responsible for turning what they learn into something repeatable.

Vishvesh and Parth have perfected that process with our Customer Success team, training and enabling our field engineers to run the same mapping-and-modeling work at the next plant, and the one after that. Microsoft and Litmus Customer Success, in turn, train our shared certified systems integration partners to roll out the same approach at scale, across dozens, sometimes hundreds, of facilities, without needing a Vishvesh or a Parth on-site for every one of them. It’s the same scaling logic Microsoft is applying with Frontier Company, which works alongside global systems integrators - embedded experts prove the pattern and a trained partner ecosystem carries it everywhere else.

The virtuous circle actually forms at that last step. When Vishvesh, Parth, or one of their trained field engineers walks into a customer’s plant, the customer’s proprietary process knowledge (the tribal know-how, which alarm code means what on which line, the industrial IP that makes that plant run) stays exactly where it belongs: with the customer. What travels back to us is something narrower and more generalizable: the pattern of the connector logic, the data model, and the schema approach that turns out to work not just for one plant but for a whole class of them.

This is the same line Microsoft draws in its Frontier commitments. The idea is that a customer’s intelligence should compound from within. For a manufacturer, that intelligence is decades of process knowledge embedded in how each line runs. A well-modeled edge data layer amplifies it, making it usable by Azure AI across every plant, without ever handing it away.

Patterns and data intelligence like this are what become joint solution. Litmus Edge Manager’s Site Model, the very capability Microsoft just certified for Manufacturing AI, exists because field-tested naming-and-mapping work like this gets folded back into the roadmap rather than left in a project folder. To us, Microsoft’s certification is independent confirmation that the pattern really did generalize. And because Microsoft’s MVP program recognizes the people who did that field work separately from its product-certification process, we end up with two independent, cross-checking signals of trust arriving in the same season: one for the product, one for the people who build what the product eventually becomes.

This combination keeps the circle turning. A certified product and nationally recognized MVPs make it easier for Microsoft’s field teams to co-sell Litmus into new manufacturing accounts. New accounts bring new plant-specific problems. New problems mean more of exactly this kind of naming-and-mapping work for Vishvesh, Parth, and the rest of the team. More engagements mean more patterns worth extracting, and more candidates for the next product certification. The product and the people keep feeding each other, and Microsoft’s recognition system validates both sides independently, which is why the circle keeps expanding instead of stalling out.

Vishvesh will be at the upcoming Microsoft Partner Advisory Council, representing Litmus and bringing the field-level perspective that earned MVP recognition. Vishvesh be sharing what embedded engineering looks like on real plant floors, how the protect-and-amplify principles at the core of Microsoft’s Frontier vision translate to industrial data and IP, and what manufacturers need from the Azure IoT Operations and Manufacturing AI ecosystem next. If you’ll be attending, whether you’re a Microsoft team member, a systems integration partner, or a fellow ISV, we’d encourage you have a chat with Vishvesh or Chris Hilderbrand, Head of Cloud and AI Partnerships. They’re always glad to trade notes on the hard problems.

Microsoft’s own criteria make an important distinction here: the MAICPP certified software designation is a product and commercial bar, not an individual credential, while the MVP program is Microsoft’s separate, people-side recognition. Earning both in the same season isn’t a coincidence of timing, in our view. It’s evidence that solving unglamorous, plant-specific problems, like getting two systems to agree on what “Alarm_12” means, is what earns both kinds of recognition at once: certifying what we ship and getting recognized for who shows up to make it work.

For manufacturing teams sizing up their next edge-to-cloud project, that’s what these announcements say together. The software clears Microsoft’s technical bar for Manufacturing AI. The people who will show up to make it work on your specific plant floor are, by Microsoft’s own count, among a handful of the best in the country at exactly that job. And the way we work, embedded with your team, focused on measurable outcomes, and protective of the process knowledge that makes your plants yours, is the same model Microsoft is now betting on at global scale with Frontier Company.

Congratulations again to Vishvesh and Parth, and thank you to the customers who handed us the hard problems that made this recognition possible. If your team is wrestling with its own edge-to-cloud connection problem, we’d like to hear about it. Reach out to [email protected] to schedule a meeting.

It is a Microsoft certification for partner solutions that show real AI capability on Azure for manufacturing. Solutions must pass a technical audit, meet customer-success thresholds and submit manufacturing-specific case studies. Litmus Edge Manager's Site Model earned it in 2026.

It standardizes and maps industrial data across plants. Equipment and events that different integrators named differently over many years, such as "Alarm_12" and "E-Stop-Spindle" for the same spindle fault, are aligned into one consistent model, so manufacturers can run cross-site analytics and Azure AI.

Litmus Edge connects to Azure IoT Operations through the Akri Litmus Connector. Manufacturers can then manage edge data securely using Azure Arc, Azure Resource Manager and Microsoft Entra ID, with real-time device and asset discovery. The partnership was announced at Hannover Messe in March 2025 and expanded in November 2025.

They are Litmus engineers whom Microsoft named Most Valuable Professionals for the second year in a row, two of five U.S.-based MVPs in the IoT category. They earned it through two years of field work with Microsoft building the Litmus Edge Bridge for Azure and Azure IoT Operations.

Microsoft Frontier Company is a $2.5 billion Microsoft initiative that brings more than 6,000 experts into customer organizations to deploy AI. Litmus's model works the same way: embedded engineers who are accountable for measurable outcomes, a trained partner ecosystem that scales the work, and protection of the customer's proprietary knowledge.

Yes. The plant-specific process know-how stays with the customer. Only generalizable patterns, such as connector logic, data models and schema approaches, go back into Litmus's product.