The Future of Coding: AI Agents & the Next Tech Revolution // Ricky Doar

Why It Matters

Mastering AI prompting and context management transforms coding efficiency, letting firms modernize legacy systems faster while mitigating the risks of over‑reliance on autonomous code generation.

Key Takeaways

• Effective AI prompting requires breaking problems into bite‑sized tasks.
• Maintain minimal, relevant context; avoid overloading the model’s window.
• Use semantic code indexing for brownfield projects, not just greenfield.
• Avoid over‑reliance on AI; retain strategic architectural decision‑making.
• Leverage plan mode and markdown artifacts to reset context cleanly.

Summary

The conversation centers on Cursor, an AI‑driven coding assistant, and how developers are adapting to a new paradigm where large language models act as pair programmers or autonomous agents. Ricky Doar and his guest discuss the rapid adoption of Cursor across enterprises, the recent funding round, and the broader shift toward AI‑augmented software development.

Key insights include the emergence of a distinct skill set: engineers must learn to decompose complex tasks into bite‑sized prompts that fit within model context windows. Context management is critical—overloading the window leads to hallucinations, so starting fresh chat sessions or using plan‑mode markdown files is recommended. Cursor’s semantic indexing of entire codebases (even half‑million‑file repositories) enables effective brownfield work, debunking the myth that AI excels only in greenfield projects. Model upgrades (Claude 3.5→4, Gemini, GPT‑4.5) continually expand what can be automated.

Notable examples illustrate best practices: "The best people break large problems into bite‑sized tasks that the AI can actually accomplish," and the team highlights Cursor’s ability to search 500,000 files instantly to add authentication or follow existing design systems. They also warn against over‑reliance, noting that AI should not replace strategic architectural decisions, and that excessive context can cause the model to make arbitrary, error‑prone choices.

The implications are clear: developers must treat AI as a collaborative tool, mastering prompt engineering and context hygiene to reap productivity gains while avoiding complacency. Enterprises that embed these practices can accelerate feature delivery, modernize legacy codebases, and stay competitive in the emerging AI‑first software development era.

Original Description

March 3rd, Computer History Museum CODING AGENTS CONFERENCE, come join us while there are still tickets left.

https://luma.com/codingagents

Thanks to @ProsusGroup for collaborating on the Agents in Production Virtual Conference 2025.

In this session, Ricky Doar, VP of Solutions at Cursor, shares actionable insights from leading large-scale AI developer tool implementations at the world’s top enterprises. Drawing on field experience with organizations at the forefront of transformation, Ricky highlights key best practices, observed power-user patterns, and deployment strategies that maximize value and ensure smooth rollout. Learn what distinguishes high-performing teams, how tailored onboarding accelerates adoption, and which support resources matter most for driving enterprise-wide success.

Bio //

Ricky Doar is the VP of Solutions at Cursor, where he is building the pre- and post-sales Solutions organization for the AI-powered code editor. A seasoned product and technical leader with over a decade of experience in developer tools and data platforms, Ricky previously served as VP of Field Engineering at Vercel, where he led global technical solutions for the company's next-generation frontend platform.

Prior to Vercel, Ricky held multiple leadership roles at Segment (acquired by Twilio), including Director of Product Management for Twilio Engage, Group Product Manager for Personas, and RVP of Solutions Engineering for the West and APAC regions. He also worked as a Product Engineer and Senior Sales Engineer at Mixpanel, bringing deep technical expertise to customer-facing roles.

A Prosus | MLOps Community Production