A recent MIT study showed that 95% of AI projects fail to deliver value, but there is an exciting class of tools that solve some of the key challenges, they’re called Visual Agentic AI Tools (VAAT) and they make AI fast and easy! As a result the market for these tools is exploding with new tools, massive VC investments, and multi-billion dollar acquisitions by the largest software companies who are eager to get in on the gold rush for these low-code/no-code (LCNC) tools. All of this means you can build amazing AI agents rapidly with zero knowledge of coding. In fact, Salesforce found that these tools are so good that even the code jockeys are using them to build agents much faster.
When venture capitalists see technical innovations that unlock whole new industries they pounce, throwing big money at entrepreneurs. The VAAT market was fresh meat to VCs about 18-months ago and now there are more than 27 VAAT tools making agent building a snap; you actually snap components together to create them. Powering all of this are a few technical innovations: (1) a visual tool for easily wiring components (React Flow); (2) AI improvements; and (3) Tools for connecting agents to data sources/functions (MCP and RAG). This combination of tools has created a wave of innovation that analysts expect will grow at 45.8% per year (Grandview Research). Soon, many non-programmers will be using these tools to build agents. More importantly, a majority of business users will be using these agents with many not even knowing they are using them.
Traditional IT focused on buying best-of-breed applications, like Salesforce for CRM and Workday for HR, and making them work together. While you could customize these apps somewhat, you would basically modify your workflows to their app. Agentic AI breaks through these silos, enabling you to create your own workflows that weave them together the way you want them to work. This is reshaping the IT industry and making a lot of companies very nervous because it weakens their walled gardens (ecosystem of tools, apps, partners, and consultants) which weakens vendor lock-in. Of course, this results in more competition from upstarts. This explains why the big boys have been on a tear to acquire these tools and integrate them into their ecosystems to extend their walled garden into the Agentic AI Age. So far in 2025, these are the acquisitions:
| Acquirer | Target | Deal Value | Date |
| ServiceNow | Moveworks | $2.85B | March 2025 |
| IBM | DataStax | ~$3B | February 2025 (closed May) |
| Workato | DeepConverse | Undisclosed | March 2025 |
| Workday | Flowise | Undisclosed | August 2025 |
| Salesforce | Regrello | Undisclosed | August 2025 |
The play is to optimize their respective VAATs for their ecosystem, while extending its reach to other apps as well. Some of the elephants are buying, while others build, and still others integrate open source tools. The platform companies like Microsoft, Google, and SAP are in the game, but Meta, Apple and Oracle not so much yet. Azure and Google Cloud (GCP) are listed above, but AWS is also leaning-in from a cloud perspective. I expect more engagement from pure-play AI companies like XAI, OpenAI, Anthropic, etc. All of them will be exposing their apps to this massive market using both MCP and custom components that go right into the drag-and-drop palettes of these VAATs.
These tools enable you to drag out components (function-specific elements), draw lines between them (think inputs and outputs), select a few options and you’ve built an agent. You can then make copies of this agent (called blueprints, templates or flows) and modify it to suit your specific needs. Some of the VAAT vendors provide collections of templates for download on Github to make the process of agent building even faster and less prone to errors.
Why LCNC? It slashes development time by 70-80%, per Gartner's 2025 Magic Quadrant for Enterprise Low-Code Application Platforms. Salesforce echoes this: 85% of developers building agentic AI lean on LCNC tools, with 77% saying it democratizes AI dev and 78% noting it scales projects faster. This isn't hype; it's hyperautomation in action.
To give you a lay of the land, here's a table of key players, with estimated active users as of September 2025 (based on downloads, customers, and growth metrics). Total across all? A whopping ~52M users, but watch for overlaps.
| Tool | Estimated Users (as of September 2025) |
| Microsoft Power Automate | 30,000,000 |
| ServiceNow Workflow Automation | 10,000,000 |
| Make.com | 5,000,000 |
| Zapier | 5,000,000 (est.) |
| UiPath | 3,000,000 |
| Botpress | 1,000,000 |
| Apache Airflow | 1,000,000 |
| Google Vertex AI Agent Builder | 1,000,000 |
| n8n | 230,000 |
| Dify | 200,000 |
| ActivePieces | 100,000 |
| CrewAI | 100,000 |
| Flowise | 100,000 |
| Kore.ai | 100,000 |
| Stack AI | 90,000 |
| Langflow | 50,000 |
| Relevance AI | 50,000 |
| Salesforce Agentforce | 50,000 |
| Haystack by deepset | 20,000 |
| Lyzr | 20,000 |
| Lindy | 10,000 |
| Prefect | 10,000 |
| Eden AI | 5,000 |
| Synthflow | 5,000 |
| Vellum | 5,000 |
| Airia | 1,000 |
| Sim | 1,000 |
Sources: AI aggregated from GitHub stars, revenue indicators, and reports.
The Visual Agentic AI Tools (VAATs) enable programmers and non-programmers to build powerful agents that execute personalized workflows across various data sources and applications. By enabling non-programmers to use these tools, the base of potential agent builders is huge! Reuse of templates makes it even easier. This explains why this market is an absolute rocketship. Get in now and start learning, these skills will be in high-demand as agents reshape the IT landscape.
Stay tuned for the second post in this 4-part series
Subscribe to our newsletter to receive the latest updates and promotions from MPH straight to your inbox.
