- brian.finance
- Posts
- Does UiPath's Maestro Have Product Market Fit? Here's What The Data Tells Us
Does UiPath's Maestro Have Product Market Fit? Here's What The Data Tells Us
Analysis of Fortune 100, 500, and 2000 companies shows three distinct AI strategies and where UiPath Maestro fits
Brian McCormick
December 03, 2025
Where is UiPath and their new orchestration product, Maestro, in terms of product market fit?
In this article, we will take a look at a random sample of 75 Fortune 2000 companies to see how they are adopting AI and if Maestro is a good match.
Before we begin, let’s briefly go over what UiPath Maestro is and who it is for.
Maestro is a conductor for the AI workforce. In a modern enterprise, work is no longer done solely by a rules based bot or a human. Work is handed back and forth between rules based robots, AI, and human validators.
Maestro is the one in charge of organizing the process flow between these systems. It is the “orchestration layer”.
In this orchestration product, you can plug and play your own or third party AI agents to work together in process flows.
Watch an example here:
The Three Categories of AI Adoption Today
We are in a period of fragmentation where multiple strategies are being explored. The three major categories for AI solutions are as follows:
1. Data Focused Automation
These are agents that live where the data lives.
This is Salesforce Agentforce, Microsoft Copilot
The business incentive is context & speed.
Why they win: If an agent needs to read a CRM history and email a customer, Salesforce does it instantly. Moving that data to third parties adds friction.
The risk: Their weakness is cost & lock-in. If you use Salesforce Agentforce, you are paying Salesforce pricing for every token and interaction.
2. Process Focused Automation
These are agents that manage the communication between systems.
This is UiPath Maestro, Camunda, ServiceNow
The business incentive is reliability and breadth of automations
Process orchestrators win here because they reliably bridge modern AI with old tools that have no APIs. Products like Maestro can connect automations from different systems.
The risk: Vendors that exist where the data is, such as Salesforce may build sufficient automation workflows for their specific areas, so clients won’t expend the energy to go elsewhere.
3. Proprietary In-House Automation
These are in house solutions. These are bespoke and built by a business to orchestrate their own solutions.
Examples include Lemonade’s Blender, Klarna’s AI assistant, and others.
The business incentive is maintaining data control and competitive differentiation
If a hedge fund is building a trading agent, they build it in-house to ensure they own the architecture and to prevent their secret sauce from being locked inside a competitor's ecosystem.
The risk: You are trading vendor lock-in for maintenance debt. By building it yourself, you force expensive internal engineers to maintain plumbing forever instead of building new product features, while third party businesses update their solutions automatically over time.
These are the three main categories of AI automation deployment being offered and used by businesses. So what does adoption look like today and which is winning?
Let’s take a look. I will breakdown what is being deployed by a random sample of Fortune 100 companies, Fortune 500, and Fortune 2000 companies. As the solutions of different sized enterprises may vary.
(Skip ahead to the chart summary if you don’t want to read the breakdown by business)
Fortune 100 Automation Adoption by Category
Random sample analysis of 25 businesses AI adoption by type
Company | Primary Strategy | Specific Solutions Used | Notes / Evidence |
Walmart | Proprietary In-House | "Sparky" (Negotiation), "Marty" (Store Ops), "Element" Platform | Built internal agents using custom models to own the data. Treats AI architecture as a trade secret to dominate retail margins. |
Amazon | Proprietary In-House | AWS Bedrock Agents, Olympus, Titan | As a cloud provider, they build their own "Sovereign Agents" to sell the resulting tech to others. |
Apple | Proprietary In-House | Apple Intelligence, Siri, OpenAI (Plugin) | Orchestration is handled on-device by a custom OS layer. OpenAI is just a "tool" called by the Apple orchestrator. |
UnitedHealth Group | Process Focused | "Smart Choice" (Custom), Microsoft Copilot, Optum | Heavily utilizes Azure for data, but relies on deep operational orchestration for claims processing logic. |
Berkshire Hathaway | No Information | N/A | Strategy is decentralized across subs (GEICO, BNSF). No public "Agentic" standard announced. |
CVS Health | Process Focused | UiPath, Microsoft Azure AI, "Aetna" Voice Agent | Massive user of UiPath for backend claims. Uses voice AI to handle prescription calls. |
McKesson | Process Focused | SAP, Specialized Supply Chain Tools | Heavy reliance on ERP-centric orchestration tools. Focus is logistics reliability over building custom "AI Brains." |
Alphabet (Google) | Proprietary In-House | Gemini, Vertex AI Agent Builder | Utilizes their own Gemini models for all internal orchestration. They are the definition of "Build." |
Cencora | Process Focused | World Courier Tools, Legacy ERP | Focused on temperature-controlled supply chain logistics. Orchestration manages physical goods across borders. |
Microsoft | Data Focused | Copilot Studio, Azure AI | They run the company on their own stack (Dogfooding) to prove Data Gravity is sufficient for enterprise. |
Costco Wholesale | No Information | N/A | Notoriously low-tech in operations. Likely legacy-based with no public "Agentic" strategy. |
Cigna Group | Proprietary In-House | Evernorth Custom Models | Builds custom predictive agents for patient care, viewing the logic as core IP. |
AT&T | Proprietary In-House | "Ask AT&T" (Built on Azure OpenAI) | Runs on Microsoft's cloud, but the orchestration layer was custom-built to secure code and HR data. |
Cardinal Health | Process Focused | SAP, Legacy RPA | Low-margin distribution requires high-reliability process automation rather than experimental GenAI. |
Chevron | Proprietary In-House | "Digital Twin" Agents, Microsoft Copilot | Uses custom physics-based agents for drilling; standard Copilot for office workers. |
Home Depot | Proprietary In-House | "Sidekick" App, Glean, Google Cloud | Built "Sidekick" to orchestrate inventory tasks for employees. Heavy user of Glean for internal search. |
General Motors | Proprietary In-House | "OnStar", Google Cloud | Recently dropped Apple CarPlay to build their own in-car orchestration OS to own the customer data. |
Ford Motor | Proprietary In-House | Latitude AI, Internal Tools | Created a subsidiary to own the orchestration of autonomous driving and manufacturing logistics. |
Elevance Health | Proprietary In-House | "Health OS" | Attempting to platformize their own data with a custom OS rather than relying solely on external vendors. |
JPMorgan Chase | Proprietary In-House | "LLM Suite", "COIN" | Explicitly blocked public ChatGPT. Built internal platform "LLM Suite" to ensure regulatory compliance. |
Verizon | Process Focused | GenAI Contact Center, Legacy Stack | Large deployment of GenAI for Customer Service routing, connecting the call center stack to modern AI. |
Kroger | Proprietary In-House | 84.51° (Data Science Subsidiary) | Uses their internal tech subsidiary to build custom inventory and customer personalization agents. |
Fannie Mae | Process Focused | IDP Tools, Compliance Orchestrators | Highly regulated. Uses orchestration to ensure mortgage compliance and document processing. |
Bank of America | Proprietary In-House | "Erica", Internal Coding Agents | Early pioneer of the custom financial assistant. Continues to build the architecture in-house. |
Comcast | Proprietary In-House | "Xfinity Assistant" | Controls the hardware (routers) and the software agents that manage them. |
Here’s the overall breakdown of the sample:
Strategy Category | Company Count | Percentage |
Proprietary In-House | 15 | 60% |
Process Focused | 7 | 28% |
No Information | 2 | 8% |
Data Focused | 1 | 4% |
Total | 25 | 100% |
Fortune 500 Automation Adoption by Category
Random sample analysis of 25 businesses AI adoption by type
Company | Primary Strategy | Specific Solutions Used | Notes / Evidence |
Starbucks | Proprietary In-House | "Deep Brew" | Built a custom AI platform ("Deep Brew") to orchestrate personalized offers and inventory. Recently shifted to "Human-Centric" AI to assist baristas, not replace them. |
Bristol-Myers Squibb | Proprietary In-House | Internal Research Agents | Uses custom AI for drug discovery and clinical trial design. They view this logic as core IP and do not outsource the "science brain" to generic platforms. |
US Foods | Process Focused | Descartes, "Pronto" | Uses specialized routing and logistics orchestration (Descartes) to manage supply chain exceptions. Built "Pronto" for small-truck delivery logic. |
Salesforce | Data Focused | Agentforce | They are the platform. They run their own sales and service operations on Agentforce to prove the model works. |
Thermo Fisher Scientific | Proprietary In-House | "Connected Lab" | Orchestrates laboratory equipment and data. This requires deep, physics-based integration that off-the-shelf IT agents cannot handle. |
Qualcomm | Proprietary In-House | Qualcomm AI Stack | Builds the actual chips and software stack for on-device agents. They are building the infrastructure for others to run agents. |
Netflix | Proprietary In-House | Machine Learning Platform (Internal) | Their recommendation engine is the agent. It orchestrates content delivery and personalization using 100% custom code. |
Paramount Global | Process Focused | ReelMind, Content Workflows | Uses specific media orchestration tools for video tagging and content distribution. Less about "GenAI Agents" and more about media supply chain. |
Capital One | Proprietary In-House | "Chat Concierge", Internal Fraud Models | Famous for being "software-first." Built their own multi-agent conversational system and fraud detection to own the banking logic. |
Mastercard | Proprietary In-House | "Decision Intelligence" | Built a custom decisioning engine that scores transactions in milliseconds. This real-time orchestration cannot be outsourced to a slow chatbot. |
Visa | Proprietary In-House | Custom Fraud Agents | Similar to Mastercard. Their "Agent" is a real-time fraud detection system built on proprietary transaction data. |
McDonald's | No Information (Reset) | Google Cloud Partnership | Recently ended their partnership with IBM for drive-thru voice agents. Currently resetting strategy, likely moving to a new partner (Google). |
Union Pacific | Process Focused | Specialized Rail Logistics | Orchestration is physical (train scheduling). Relies on heavy industrial IoT and legacy logistics platforms, not office-based AI agents. |
Southwest Airlines | Process Focused | Modernization Tools, AWS | Currently struggling with legacy tech debt. Focus is on modernizing the scheduler (Process) rather than deploying flashy AI agents. |
Delta Air Lines | Proprietary In-House | "Delta Concierge" | Built a custom agent inside their app for travelers. Orchestrates gate changes, bag tracking, and re-booking using internal data. |
Northrop Grumman | Proprietary In-House | Defense Systems | Builds autonomous agents for warfare (drones/cyber). This is the ultimate "Proprietary" category; zero chance of using public SaaS agents. |
3M | Proprietary In-House | Material Science Models | Uses custom AI to discover new adhesives and materials. The "Agent" here is a chemist, not a chatbot. |
General Mills | Process Focused | Supply Chain Control Tower | Uses AI routing engines to optimize truckloads. The focus is purely on logistics efficiency and cost reduction. |
Gilead Sciences | Proprietary In-House | "GEMS" (via Genesis) | Partners with Genesis Therapeutics to use custom generative AI for drug discovery. The orchestration is molecular, not administrative. |
BlackRock | Proprietary In-House | "Aladdin Copilot" | Built a custom copilot for their "Aladdin" investment platform. It is a "Sovereign Agent" designed to give traders an edge. |
PayPal | Proprietary In-House | Fraud Detection Agents | Uses graph-based agents to detect fraud rings. The logic is proprietary and central to their margin. |
AutoZone | Process Focused | Inventory Management Systems | Relies on EDI and supply chain orchestration to keep parts in stock. The "Agent" is a re-ordering algorithm. |
Sherwin-Williams | Proprietary In-House | "Color Expert" App | Built a custom computer-vision agent for customers to match paint colors. Internal R&D uses custom formulation agents. |
Kimberly-Clark | Process Focused | NuvoOS (Logistics) | Partnered with Nuvocargo to orchestrate cross-border logistics. Focus is on shipping efficiency. |
Colgate-Palmolive | Proprietary In-House | "Hello" Brand AI | Using AI as a "Strategic Amplifier" for marketing. Integrates AI into product development and consumer insights. |
Here’s the overall breakdown of the sample:
Strategy Category | Company Count | Percentage |
Proprietary In-House | 16 | 64% |
Process Focused | 8 | 32% |
Data Focused | 1 | 4% |
Total | 25 | 100% |
Fortune 2000 Automation Adoption by Category
Random sample analysis of 25 businesses AI adoption by type
Company | Primary Strategy | Specific Solutions Used | Notes / Evidence |
Harley-Davidson | Process Focused | "Manufacturing 4.0", Siemens/PTC | Focus is on the "Smart Factory." Orchestration handles assembly line data and predictive maintenance for machinery. |
Mattel | Proprietary In-House | OpenAI Partnership, DALL-E Integration | Partnered with OpenAI to build a custom "Idea Engine" for toy design. They treat creativity as their core IP. |
Hasbro | Process Focused | Supply Chain Analytics, Legacy ERP | heavily focused on "Blueprint 2.0" to cut costs via supply chain orchestration. (Wizards of the Coast uses digital AI, but the mothership is focused on Ops). |
NY Times | Proprietary In-House | "Echo", Custom Newsroom Tools | Built "Echo" (internal) to query their archives. Explicitly avoids using public tools that would train on their data without permission. |
Wendy's | Proprietary In-House | "FreshAI" (Google Cloud) | Built a custom drive-thru agent on top of Google's LLMs. It is not a standard "chatbot"; it integrates with the kitchen display system (KDS). |
Crocs | Process Focused | Warehouse Automation, SAP | Their massive growth required a supply chain overhaul. Automation is focused on moving boxes and inventory visibility. |
Columbia Sportswear | Process Focused | Microsoft Dynamics 365, Supply Chain | Recently overhauled their ERP. The focus is on orchestrating the flow of goods, not building flashy customer agents. |
Dropbox | Proprietary In-House | "Dropbox Dash" | Built their own "Universal Search" agent. They are trying to be the Data Gravity platform for files, competing with Microsoft. |
Zoom | Proprietary In-House | "AI Companion" | Like Salesforce, they are building the agent into their product to prevent users from leaving. It orchestrates meetings, docs, and chat. |
Yelp | Proprietary In-House | "Yelp Assistant", "Yelp Guest Manager" | Built a custom consumer agent to find plumbers/restaurants. It orchestrates the connection between user intent and booking APIs. |
Zillow Group | Proprietary In-House | "Housing Super App", Neural Search | The "Zestimate" was the original agent. Now building a "Super App" agent to orchestrate the entire home buying transaction (Loans + Agents). |
TripAdvisor | Proprietary In-House | "Trips" (OpenAI powered) | Built a custom itinerary builder. Orchestrates hotel data, reviews, and maps into a single travel plan. |
Levi Strauss & Co. | Proprietary In-House | "Project Stitch", "AL" (Bootcamp) | Created "Stitch" (Custom Store OS) to help stylists find sizes. View their data as a "fashion moat." |
Under Armour | Process Focused | Supply Chain Optimization | heavily focused on inventory efficiency ("Project 2025"). Orchestration is about getting the right shirt to the right store. |
American Eagle | Process Focused | "Quiet Platforms" | Acquired a logistics company to build their own supply chain orchestration network. They sell this "process" to other brands. |
Wyndham Hotels | Process Focused | "Wyndham Connect", Canary Technologies | Rolled out "AI Agents" for guest calls, but bought the tech from Canary/Bandwidth rather than building a custom LLM. |
Six Flags | Process Focused | Operational Efficiency Tools | Focus is on "Guest Flow" and dynamic pricing orchestration. |
Domino's Pizza | Proprietary In-House | Voice AI, Delivery Algorithms | They call themselves a "Tech company that sells pizza." Built custom voice agents and routing logic (Core IP). |
Planet Fitness | Process Focused | Salesforce Marketing Cloud, Retention AI | Uses orchestration to identify "at-risk" members and trigger retention offers. It's a classic "CRM Orchestration" play. |
Roku | Proprietary In-House | Content Recommendation Engine | Their home screen is the agent. It orchestrates ad delivery and content suggestions across all streaming apps. |
Snap Inc. | Proprietary In-House | "My AI" | One of the first to launch a custom agent. It is core to their engagement strategy, not a back-office tool. |
Proprietary In-House | Computer Vision Agents | Their "Visual Discovery" engine is a proprietary agent that understands style and intent without text. | |
DraftKings | Proprietary In-House | Betting Engine, Anti-Fraud AI | The "Oddsmaker" is now an agent. They build custom risk models to set lines in real-time. |
Etsy | Proprietary In-House | "Gift Mode", Graph Search | Built a "Knowledge Graph" agent to understand "presents for dad" rather than just keyword matching. |
Take-Two | Proprietary In-House | Zynga Ops, Game Engines | While skeptical of "GenAI Games," they use proprietary agents for mobile user acquisition and game testing (Zynga). |
Here’s the overall breakdown of the sample:
Strategy Category | Company Count | Percentage |
Proprietary In-House | 16 | 64% |
Process Focused | 9 | 36% |
Data Focused | 0 | 0% |
Total | 25 | 100% |
Breakdown Analysis of Fortune 100, 500, and 2000 Automation Adoption
The primary strategy of companies by type has a clear pattern. Fortune 100, 500, and 2000 companies are currently adopting solutions at roughly the same rate. Here is a chart comparing adoption of each type as their primary automation choice.
As we can see, the overwhelming primary strategy of choice is proprietary in-house. It’s consistent across the board at around 60-70%.
Next we have process focused adoption. This is around 1/3rd of businesses. It also has slightly higher adoption the smaller the enterprise business is.
Finally, data focused as primary adoption is the lowest, and is not the primary use case for nearly every business.
Why Proprietary In-House Automation is in 1st Place
Our analysis of the Fortune 100, 500, and 2000 reveals that the majority of large enterprises are building, not buying, their primary AI orchestration.
Why are 60% of companies taking on the maintenance debt of building their own agents rather than just renting Salesforce Agentforce or Microsoft Copilot? The answer comes down to three factors: differentiation, sovereignty, and cost.
For companies like JPMorgan Chase, Walmart, or Netflix, the AI agent is the product itself.
If JPMorgan uses an off-the-shelf banking agent from a public vendor, they have the exact same capabilities as Wells Fargo.
To achieve competitive advantage, they must own the architecture of how the agent thinks, weighs risks, and makes decisions. They view the orchestration layer as core IP, and you do not outsource Core IP.
Why not have your own proprietary agent work inside a third process orchestrator? If you plug your custom agent into Maestro, you own the worker, but UiPath owns the manager. The risk is your secret sauce is also the complex decision tree of how that agent interacts with the world.
If you build that workflow inside Maestro’s proprietary .xaml or BPMN charts, your business logic is trapped in UiPath's file format and you still pay UiPath a management fee for using your own agent.
Why Process Orchestration is a Strong Runner-Up
This sector is not winning the tech focused companies, but it is winning the physical and regulated businesses in healthcare, logistics, manufacturing, and insurance.
These businesses often have legacy tech debt and run on old legacy infrastructure. Critical data is trapped in mainframes and old on-premise ERPs, with applications that don’t have modern APIs.
A Salesforce agent cannot easily see or click a button inside a 1990s mainframe application.
This is UiPath’s strength. Maestro wins as part of this category because it acts as a bridge. It can have a modern LLM read an email and then command a legacy robot to update the mainframe. For companies like CVS or Union Specific, this provides the ability to bridge old and new technology.
So why choose a process orchestration business over doing it in house?
Building an AI agent can be somewhat easy. Building the system that orchestrates an agent is harder.
Most operational workflows are not completed instantly. It requires the system to ingest data, wait for a human review, retry an API call if the server is down, and wake up only when a specific vendor replies.
If you build this in custom code, you have to engineer the orchestration layer from scratch.
Platforms like UiPath Maestro can do orchestration out of the box. They handle the waiting and retries automatically. Enterprises buy them so they don't have to engineer the concept of time. They just have to engineer the what and why.
Data Focused - Why They Are Losing As Primary Automation Layer
You might look at the breakdown chart and think, "Wait, doesn't everyone use Microsoft Copilot?" They do. But there is a difference between a productivity tool and a competitive advantage. Businesses are using these products, but they are NOT the primary strategy.
The reason data-focused automation shows up as the primary strategy for only ~4% of companies is likely due to these four factors:
They are commodity products. Microsoft Copilot and Salesforce Agentforce don't differentiate the business. Everyone can have them.
Using Copilot to summarize emails is helpful, but it doesn't give Walmart a competitive edge over Amazon.
Companies use these tools for standard work (HR, IT, Email), but when asked about their primary AI strategy (the thing that drives revenue), they point to their proprietary builds.
CFOs are wary of the pricing models for data-focused platforms, which often charge by the token.
If a logistics company needs to process 1 million shipping updates a day, running that through a Salesforce agent that charges per interaction is financially unpalatable.
They prefer process or proprietary solutions where they pay for the fixed infrastructure rather than a tax on every single action.
Does Maestro Have Product Market Fit Moving Forward?
Our consistent sample data shows that Maestro’s orchestration layer has product market fit. It’s type of solution is being deployed by the 30-40% well defined segment of the market we see in Fortune 100, 500, and 2000 companies.
However, 2/3rds of businesses are still choosing to focus on proprietary orchestration, which means Maestro is unlikely to become a universal standard.
While proprietary in house orchestration will create advantages for new businesses, legacy businesses may find Maestro the perfect fit.
Companies in these industries have pains that Maestro can solve:
High legacy tech debt
Long-running processes with wait times
Require an audit trail solution
Maestro has product market fit for the enterprises that cannot escape its legacy infrastructure. In a world where everyone is rushing to build smart agents, UiPath has smartly positioned itself as a reliable solution for legacy businesses looking to connect their business together.
However, there are threats to UiPath.
Right now, Maestro wins because it bridges AI to Legacy. But if Salesforce or Microsoft builds better legacy connectors, or if Legacy apps finally build real APIs, Maestro's value proposition shrinks.
Maestro is unlikely to be the universal standard that runs every aspect of a business. Proprietary builds are winning that war. However, UiPath has already secured the unglamorous layer of the economy that requires reliability.
For the 30-40% of the enterprise market that relies on complex, physical, or regulated operations, Maestro is positioned as a necessary bridge between the old world and the new.
If you enjoyed this article, you may also be interested in reading Jim Cramer's 10 Failed Dot Com Bubble Stocks & Why They Failed.
Our content is provided subject to our Disclaimer.
Reply