AI agents as 'first-class citizens' in software development

"Developer experience" is evolving into "agent experience" as MCP elevates AI agents into primary users and direct consumers of developer tools

Approximately six months ago, I wrote about how the AI revolution is catalyzing an evolution in the developer economy, driving software development towards unprecedented levels of abstraction and democratization. In the months since, advancements such as vibe coding have pushed these concepts to new heights:

Source: Post from @karpathy

But with the pace of innovation progressing at such rapid speed, the impact of AI on the developer landscape hasn’t stopped there. At the end of 2024, Anthropic (a Bessemer portfolio company) open-sourced Model Context Protocol (MCP), a new standard that enables developers to build secure, two-way connections between their data sources and AI-powered tools. MCP has been gaining buzz within the community since its launch (chart below), but it received outsized attention in the past few weeks after OpenAI announced plans to embrace the standard.

Source: Turing Post, “What Is MCP, and Why Is Everyone – Suddenly!– Talking About It?” (3/15/25)

There is much to look forward to as the industry converges toward MCP as a standard. Likened as the USB-C port of AI applications, MCP has been game-changing for developers in terms of unlocking interoperability and simplifying development. As I highlighted in Bessemer’s Data 3.0 Lakehouse Roadmap last week, interoperability is a powerful and growing phenomenon in both the data and AI landscapes.

Beyond increased efficiency and flexibility for developers, I’m most excited about the possibility of MCP elevating the status of AI agents to “first-class citizens” in the software development cycle. MCP supercharges AI agents by enabling them to complete actions as generalizable, interoperable systems that are context-aware. As Ksenia Se of Turing Post articulates:

“MCP is all about the Action part – specifically, giving agents a standardized way to perform actions involving external data or tools. It provides the plumbing that connects an AI agent to the outside world in a secure, structured manner. Without MCP (or something like it), every time an agent needs to do something in the world – whether fetching a file, querying a database, or invoking an API – developers would have to wire up a custom integration or use ad-hoc solutions. That’s like building a robot but having to custom-craft each finger to grasp different objects – tedious and not scalable.”

Blender MCP and Ableton MCP from developer Siddharth Ahuja are exciting examples of how user interactions with AI are being completely reimagined through MCP. As AI agents improve in completing actions with strong guarantees of reliability and entitlement, they are becoming direct users and consumers of developer tools. This will have profound implications on the landscape.

From agents becoming a new client type for service providers to app stores/tool discovery catered to agents, a whole new ecosystem is sprouting in preparation for a world where developer workloads shift from being predominantly executed by humans to being driven by agents:

Source: Kenneth Auchenberg (AlleyCorp), “Integrations and the service-provider side of AI and agentic systems” (Jan 2025)

Source: a16z, “A Deep Dive Into MCP and the Future of AI Tooling” (3/20/25)

“DevEx” is becoming “AEx”

AI coders are no longer subordinate co-pilots to human developers; they are now directly taking the wheel in software development. The best developer tools of tomorrow won’t just serve humans, but will need to cater for AI users as well (or as Paul Klein IV of Browserbase very aptly states “if AI can’t integrate it, ur ngmi”).

Source: Post from@pk_iv

In this shift, builders within the developer economy are already taking heed to ensure their platforms provide exceptional “agent experience” (agent-centric interfaces, documentation, deployment, efficiency, etc) and even “agent evangelism” (network effects and collaboration between agents). “AEx” is the new “DevEx” and this concept will undoubtedly shape the future of success in the developer economy during the AI age.

This post and the information presented are intended for informational purposes only. The views expressed herein are the author’s alone and do not constitute an offer to sell, or a recommendation to purchase, or a solicitation of an offer to buy, any security, nor a recommendation for any investment product or service. While certain information contained herein has been obtained from sources believed to be reliable, neither the author nor any of her employers or their affiliates have independently verified this information, and its accuracy and completeness cannot be guaranteed. Accordingly, no representation or warranty, express or implied, is made as to, and no reliance should be placed on the fairness, accuracy, timeliness or completeness of this information. The author and all employers and their affiliated persons assume no liability for this information and no obligation to update the information or analysis contained herein in the future.