Start Adopting Agentic Practices Now

Deciding where to begin with AI coding can be daunting. Human practices typically do not evolve as fast as technology, so I believe an important step is to simply start experimenting. I see a lot of commentary predicting that AI will be the end of human developers, but in my view, engineers aren’t being replaced by AI; they’re being empowered. AI coding tools should be seen, in my opinion, as productivity enhancers that can free developers from the tyranny of fiddly syntax so they can be more creative, building solutions faster, and with fewer defects. I believe the key is starting with a mindset of “there has to be a better way” and staying open to new techniques.

While basic tab-completion and casual experimentation can be a natural starting point, I advise teams to move deliberately toward fully agentic workflows, where AI handles multi-step tasks like drafting code, researching APIs, translating between languages, generating documentation, spinning up prototypes and composing test cases with minimal hand-holding. This represents a fundamental shift in how developers work, and realizing the biggest gains may take time, but I believe these are the use cases where lasting productivity improvements can emerge.

To use agentic AI effectively, I recommend that teams begin by defining small, discrete tasks. Breaking down objectives into modular components can not only help AI agents work more effectively but can also make their output easier for engineers to validate and debug. This modular approach may also facilitate running multiple agents in parallel on certain tasks like testing, documentation, code generation and analysis.

Regardless of how it’s generated, I believe it’s critical for developers to read and deeply internalize the flow of their code before committing or releasing it. While unexamined AI-generated code can heighten security risks, I’ve also seen AI-assisted security tools creating new opportunities to detect vulnerabilities faster and more thoroughly than previously possible. Teams should take secure review processes and proactive cybersecurity measures seriously, recognizing that the same technology that introduces risk can serve as a powerful counterweight when applied intentionally.

Promoting an AI-Forward Team Culture

One underappreciated benefit of AI coding tools is their ability to enforce coding standards and architectural conventions within development teams. Tools like Claude Code and Cursor enable teams to define project-level rules, preferred libraries, security requirements and design patterns in shared configuration files that a developer’s AI assistant follows. Before these tools existed, enforcing consistency typically meant relying on code reviews, onboarding documents that quickly went stale and institutional knowledge that lived in the heads of senior engineers. Now, standards can be embedded directly into the development workflow. This can be especially useful for onboarding: new joiners can follow team conventions from day one and ramp up in an unfamiliar codebase faster, with the AI serving as both a guide to existing patterns and a guardrail against common missteps.

Tools alone typically don’t change how teams work; culture can play a significant role. The companies I’ve seen get the most benefit from AI coding to date are those that intentionally socialize various tools with their software teams and create an environment that encourages experimentation and knowledge-sharing. I work with multiple companies that have made AI coding tools standard across their engineering organizations. But it doesn’t stop at a mandate. These teams also maintain AI coding Slack channels to share successful practices and help managers recognize developers who make the most of new capabilities. In addition, they hold regular AI coding meetups for enthusiastic engineers to exchange techniques, which are recorded and disseminated to the broader team.

Drive Better Outputs with Better Input

A common source of frustration with AI-assisted development I’ve witnessed is the perception that the tools don’t “get it.” But frequently, the issue isn’t with the model, it’s with the input. In my view, developers should treat an LLM like a junior engineer who can write code but lacks architectural context, product intuition and domain knowledge.

To get useful results, I encourage developers to supply clear exposition, including goals, constraints, libraries in use, performance requirements and feature definitions. Visual aids like screenshots and wireframes can be surprisingly effective, especially during debugging. The good news: LLMs typically don’t need polished writing. Bullet points, snippets or partial documentation often work just as well. Many AI coding tools now support configurations that preload relevant user documentation and standards into every prompt session, helping to ensure that outputs reflect how your team actually builds software.

AI Coding at Work: Small Shifts, Real Impact

Don’t be discouraged by a lack of immediate, measurable productivity improvements when rolling out AI coding tools. Software development is a creative process, and adapting to new workflows, especially those powered by AI, can take time. What may feel incremental at first can build toward more durable, long-term gains.

Here are four examples of where I see some of those gains starting to show:

• Faster prototyping: What once took weeks to validate can now be built, and discarded, in a matter of hours. This lowers the cost of experimentation, encourages iteration and helps teams apply lessons learned earlier in the development cycle.
• Interactive design exploration: During the design phase, developers can use LLM-powered AI chatbots to investigate implementation options far more efficiently than with traditional search. Instead of skimming documentation or scanning forums, they can ask targeted questions, such as which HTTP libraries are most commonly used in Python, and prompt the model to generate code snippets that demonstrate key differences. What once could take hours can now take minutes.
• Streamlined test development: LLMs can eliminate one of the most tedious parts of the development process: writing repetitive test cases. When seeded with clear input, output and intent, they can generate reliable tests that support CI/CD workflows, reinforce code quality and help developers stay focused on higher-leverage work.
• Tackling technical debt and legacy code: AI tools can be surprisingly effective at understanding unfamiliar or poorly documented codebases, identifying outdated patterns and generating refactored alternatives. Tasks that engineers may have avoided for months, such as migrating to a new framework, modernizing a legacy API or untangling tightly coupled modules, can become far more approachable when an AI assistant can do the heavy lifting of reading, interpreting and rewriting the code.

These improvements may not always show up in dashboards or metrics. But collectively, they represent a shift in how engineering teams can work, enabling faster validation, better decision-making and more consistent delivery.