How to Leverage AI to Enhance Developer Experience

Developer Experience (also called “DevEx” or “DX”) has a direct positive correlation with productivity and innovation.

Elsewhere, I wrote about how to spot the telltale signs of subpar developer experience.

If you want to, you can read that one here.

In that post, I suggested four ways to improve developer experience, It’s all about doing things that eliminate irritations and frustrations, and maximise joy. These were:

1. Invest in resources and processes that make life easier and nurture a sense of ownership
2. Work towards seamless collaboration to reduce data silos, improve visibility and enable more efficient workflows.
3. Optimise the development environment to reduce friction and automate repetitive tasks
4. Be conscious about culture to create a supportive, inclusive environment where people feel trusted and appreciated.

AI is ready to solve some major DX problems. 

Teams that get this right can expect to unlock substantial competitive advantages.

I want to share how I think that can work.

AI for avoiding data silos and enabling visibility

This category of tools is all about exposing high-quality information in the right places. 

That’s everything from creating useful, accurate resources to ensuring the right information reaches the right people at the right time.

For creating resources

The first use case for AI that is perhaps one of the more obvious use cases – generative AI for creating resources.

These kinds of AI can do things like write code documentation for you automatically, or summarise meetings so you don’t have to rewatch the whole thing.

Pick 1: Meeting summarisation with Otter AI. Otter AI automatically summarises meetings and captures slides to generate easy-to-digest summaries of meetings. You can connect it to your Google or Microsoft calendar and it can automatically join and record your meetings on Zoom, Microsoft Teams, and Google Meet.

Pick 2: Mintlify for automatic documentation. Almost nobody is buzzing to write docs for their code, so make the most of code-aware LLMs and get them to do it for you.

For surfacing what matters

The second use case for AI in this category is for surfacing what matters.

As far as I’m aware, there’s only one tool that observes and reflects on everything that happens across the tools you use and surfaces what matters in a bespoke, personalised way.

It’s called Stepsize AI, and I’m building it with my team.

Pick 3: Stepsize AI for surfacing what matters. You can schedule recurring reports bespoke to any team or person. Replace the daily standup, send recurring team updates, request personalised alerts when something you care about happens or generate ad-hoc reports ahead of meetings. It deeply understands the context of what your projects and goals, and can even answer questions about what’s happening. Here’s what that looks like:

Find out more about Stepsize AI.

AI for the development environment

We have to invest in reducing friction in the development environment if we want to boost productivity and satisfaction.

AI has come on leaps and bounds in the last year in developing for specific niches of the engineering process.

For a more extensive list of dev tools CTOs should know about, head here.

For code implementation

Ask any developer and they'll tell you - coding is often the stage where ideas take form but also where major bottlenecks occur. Too often, teams lose precious hours to syntax errors, library issues, and bugs that just won't squash. This hurdle can feel like a steep climb, causing unnecessary delays and frustration.

Furthermore, codebases are living, evolving organisms. Your team's work isn't just about implementing new features, but also about understanding and navigating existing code. Annotating and understanding complex codebases can be a Herculean task, especially if the original author isn't around.

AI Pick 4: Sourcegraph Cody for code implementation. It offers what feels like the next generation to GitHub Copilot, with code completion, error detection, and more that make each part of the software development process from reading, understanding and writing code much more efficient.

Honourable mentions: Codium AI for writing tests, Tabnine for an established but slightly less-cutting edge tool.

For code reviews

 it's often a time sink because it's easy to get lost in endless lines of code, particularly if you're working on a complex project or large team.

Moreover, non-technical team members often find themselves left out of the loop. When your developers, product managers, and stakeholders aren't on the same page, it can cause confusion and delay decisions.

AI Pick 5: What the Diff to improve code reviews. To optimise the review process, What The Diff is worth your time. It’s an AI assistant designed to provide summaries, review pull requests and even refactor code, making the entire review process more inclusive and efficient. It’s also great for keeping non-technical team members in the loop, and works really well with tools like CollabGPT (above) because it helps surface what matters about each pull request.

‍

Honourable mention: Planar does something somewhat similar, but with a focus on a great code review experience.

Deployment and maintenance

Let me guess. Nobody on your team lives for maintenance. But you have to deal with dependencies and migrate code. 

If only there were a way to do that hands-free.

AI Pick 6: Grit.io for hands-free tech debt management. Right on cue! Grit helps manage code migrations and dependencies, and its radar is designed to spot potential regressions, offering a smoother ride through the deployment and maintenance stage.

What you need to remember…

AI for software developers isn't a silver bullet that's going to solve all your DX issues with DX tools overnight. 

It's a facilitator. It’s about leveraging the power of artificial intelligence to bolster the creative and innovative elements of software development. 

That's where the real competitive advantage lies. 

Adapting to an AI-driven Developer Experience (DX) requires more than just selecting the right tools. It's a holistic process that impacts your team's workflows, skillsets, and overall development culture. As leaders, our roles in this transition extend beyond decision-making to driving the necessary change.

Firstly, ensure your engineering team is trained and well-equipped to work with AI tools. 

Next, reevaluate your development workflows. 

Lastly, and perhaps most importantly, nurture that culture of continuous learning and innovation. The AI landscape is continuously evolving, and to make the most of it, your team should too. Encourage your software engineering teams to stay updated with the latest AI trends, offer regular training, and create a safe space for experimentation. Remember, AI's potential is best unlocked when human creativity and machine intelligence are blended.

That’s why I created Stepsize AI, the AI Operational Intelligence Engine that takes all the manual work out of keeping informed about what matters.

By integrating with Jira, Slack, GitHub and more, Stepsize AI deeply understands the context of your projects and goals, providing bespoke, personalised updates for any person or team depending on what matters to them.

We believe Stepsize AI is a game-changer. I’d love for you to give it a spin. Try out Stepsize AI here.