In a world where AI is transforming industries across the board, Devra is leading the way in making coding more efficient and accessible for developers. It sets new standards for how software development can be streamlined and automated by allowing artificial intelligence to understand and interact with entire codebases.
In our latest episode at AllAboutAI.com, I talked with Brian Risk, the President and Founder of Devra. We discussed his journey, the inspiration behind Devra, and how this tool is making coding more efficient and accessible for developers at all levels.
Join us as we explore how Devra is reshaping the software development landscape, allowing developers to focus more on creativity and problem-solving while the AI handles the heavy coding.
Devra: Revolutionizing Coding with AI-Powered Solutions
Whether you’re an experienced software engineer looking to save time on repetitive tasks or a beginner seeking to accelerate your learning, Devra offers an AI-driven solution to make coding more intuitive and efficient.
This AI-powered software tool simplifies complex coding tasks, from understanding code structures to implementing updates. In this conversation, Brian Risk shares the inspiration behind Devra, its mission, and how it’s empowering developers to build, debug, and optimize code with unprecedented speed and precision.
Who is Brian Risk?
Brian Risk, a passionate technologist and a music composer with a diverse background in mathematics, software development, and data science, founded Devra to streamline his development workflow. His enthusiasm for AI breakthroughs like ChatGPT led him to create a tool that empowers developers to focus on innovation while automating the more tedious aspects of coding.
Let’s explore how his background in music and data science influences his work and what the future holds for Devra.
What inspired you to create Devra?
I have a non-standard background in computer science. I started in mathematics and then progressed through learning software on my own, eventually moving into the field of data science while still sticking with software development. The inspiration for developing the Devra software came from my enthusiasm and the breakthrough ChatGPT made, especially their availability of API.
When developing software, I did repetitive tasks using the ChatGPT tool. It would be best to have thorough context to get the proper output from a generalized model. The model needs to know your existing code, what frameworks you use, and the entire codebase. Doing this manually with the online tool required a lot of copying, pasting, and re-describing the project’s goals every time.
So, I created Devra, a software that hooks into the API and gives the AI agency to explore your codebase. It then asks if it can make logical updates based on the task you’re giving it. This eliminates the need for copying and pasting to get the context right and integrate the output into the codebase. It’s all streamlined within one workflow.
How do your passions for music and data science intersect in your work at Devra?
I do love music. All my friends are better musicians than me, but my strength is writing music rather than playing it. All of them are much better at performing. It isn’t easy to describe, but they think similarly to me, as though the same creative energy is used when creating software or analyzing and composing music.
As for how it intersects with data science, I feel both fields—music and software development—use the same creative parts of your mind.
What specific challenges or gaps in the market does Devra address?
The gap Devra addresses is allowing the AI agency to explore and make changes within a given codebase autonomously. It’s very flexible. It works with software and handles data files, such as CSVs or JSON objects. It can examine those and even create Python Jupyter notebooks. Essentially, it’s a general problem solver. The real gap is that Devra can work independently without overly detailed prompts, something I hadn’t seen other tools accomplish.
Have you conducted any market research before launching Devra?
In terms of market research, there wasn’t much. A few weeks before launching my product, there was an announcement for another software called DevIn, which I think many of us are familiar with. I saw a video demo, but they didn’t offer any downloadable software then, so I don’t know what level of competition it truly brings. I’ve seen other tools interface with GitHub or other repositories, but Devra fills the gap as a desktop application that interfaces with your project’s source code.
What is the philosophy behind the name “Devra”?
That’s an interesting question. Initially, I wanted a name that combined “developer” with “euphoria” to evoke a sense of joy in coding, so I thought of “Devoria.” However, that felt too long, so I shortened it to “Devra.” Then, I discovered that “Devra” is also the Hebrew word for bee, which I thought was cool. The analogy of worker bees contributing to a hive resonated with me—having a software agent working for you makes sense, and it made for a cool logo, too!
What are the main objectives of Devra?
Devra has a few primary objectives. For new projects, the goal is to go from zero to one very quickly—from a blank page to a page with usable code. That initial stage involves a lot of boilerplate tasks that have been solved repeatedly, which don’t require an expert programmer. Devra aims to add new features or quickly resolve bugs for existing projects. If there’s a bug and you get an error message, you paste the error into Devra. It analyzes which source files are affected and proposes updates. With just a click, you can apply those updates. So, the objective is to make iterations on features and updates as fast as possible.
How does Devra differ from platforms like Codota, TabNine, and GitHub Copilot?
Devra is platform-independent, unlike Copilot, which ties you into a specific IDE or repository. You can work with just text files on your computer, and the AI can still function. Devra runs on Mac, Windows, and Linux. It’s more flexible and applicable to a wide variety of coding situations. I used a prototype version of Devra to create the software itself—a Python backend with a React and Node front end. I completed both the backend and frontend in less than a month, thanks to Devra’s assistance. It helped me fly through the code and create unit tests, modules, and new views, all within a short timeframe.
Can Devra help beginners in programming or data science?
Yes. If you’re learning programming or data science, Devra can help you start projects. I’m relatively new to many of the frameworks used in Devra, like React, but with some basic programming knowledge, I quickly created a React application using the Devra prototype. Since most programming languages share core concepts, Devra helps you focus on learning the specific nuances of a new language. For beginners, it’s a fantastic way to get started without fear of making mistakes because the AI assists with the heavy lifting.
What are the key features of Devra?
Devra is relatively simple to use. First, you set up a project, such as an application you’re building or a data science analysis. Think of it as onboarding a new team member—you tell Devra the framework, the major components, what they do, the project’s objectives, how to execute unit tests, and so on. Once you do that, it remembers everything for future tasks. Then, you can create short task descriptions like “create a function to sort this list” or “create a unit test for that function” without repeating the project details. You can even paste a bug, and Devra will analyze and propose solutions based on the code.
With the rise of AI tools, do conventional tools like JetBrains or VS Code still play a significant role?
Yes, I still use conventional tools like JetBrains and VS Code. Devra fits nicely into the existing toolset. It can make some code updates, but integrated development environments (IDEs) like JetBrains are still necessary when refactoring or executing unit tests. Devra complements these tools rather than replacing them.
How does Devra ensure user data privacy and security?
Devra uses security libraries within the Python and Django environments, with robust security measures on the server side. It holds data in memory only as long as needed for the task and then purges it. In the future, I’m looking into using an Azure endpoint for a private ChatGPT instance to ensure that OpenAI doesn’t persist data.
Can you share a success story where Devra improved a client’s coding processes?
Yes, I’ve seen success with implementing new features and, particularly, with writing unit tests. Many developers dislike writing unit tests, and Devra makes it quick and easy by exploring the codebase, mocking out data structures, and implementing tests. The feedback so far has been that it makes these processes much faster and more efficient.
How does AI support software development, and why is human-led quality assurance still essential?
AI is excellent for supporting the developer role, but we still need human-led quality assurance. The AI can ensure code runs without errors, but things like user experience and correct data interpretation still require human oversight. AI can boost development efficiency, but quality assurance ensures everything works as intended.
Does Devra integrate with existing CI/CD tools like Jenkins or GitHub Actions?
Yes. I just got Devra to create a continuous integration (CI) script yesterday. I’m not well-versed in CI, but I asked it to make a CI script for running unit tests on Bitbucket pull requests, and it perfectly created the file. It doesn’t interfere with existing CI/CD workflows but helps set them up.
How do you see the role of AI evolving in the coding and software development industry?
In the next three to five years, I believe the most significant advancements will come from the models and their capabilities. Devra uses the OpenAI API backend, and they recently released a cheaper and faster model, though it isn’t as high-quality as their top-tier models.
Some tasks might require a fast, inexpensive model, while others might need a more comprehensive one. I envision incorporating model blending, which means using a suitable model for the right task.
There might be opportunities to use models from providers outside of OpenAI, like Google’s Gemini model, or even in-house models developed for specific projects. I believe the future will involve using a blend of different models to achieve the best balance of cost, speed, and quality.
How do you see AI boosting productivity without replacing jobs?
AI feels like the introduction of spreadsheets or integrated development environments (IDEs). Before spreadsheets, people did manual tabulations, and the introduction of spreadsheets revolutionized that process. Similarly, IDEs streamlined coding and debugging but didn’t reduce the number of developers. Instead, they allowed for more complex software to be built.
I see AI as an advanced form of code completion. It can double or even triple a person’s productivity, but it won’t make developers obsolete. AI gets you about 80% of the way, but there’s still a need for human creativity, logic, and the overall vision for how everything fits together. AI changes our work and increases efficiency, but it won’t eliminate jobs.
What is your top advice for entrepreneurs managing limited time and multiple responsibilities?
I’m still figuring this out myself. I’m a middle-aged entrepreneur juggling an eight-hour workday, raising children, and running Devra. For me, the most crucial advice is to stay focused. Out of a list of a hundred things you need to do, always prioritize the one most important task. Focus on executing one thing at a time.
What are the key factors to consider for success?
From my experience and observing my friends, the key to success is persistence. It’s like running a long-distance race. It would be best if you made steady progress every day. It’s not about sprinting and making quick progress in a few days only to burn out.
One of my close friends, who runs a popular YouTube channel, StatQuest, exemplifies this. He consistently made videos about statistics, building up an audience over time. He now has over a million subscribers, but it didn’t happen overnight. It’s a testament to diligence and maintaining a consistent workflow.
Key Takeaways
Brian Risk’s insights into Devra emphasize its mission to boost developer productivity by automating repetitive coding tasks. Inspired by ChatGPT’s API, Devra streamlines manual processes, allowing AI to explore and modify codebases autonomously, freeing developers to focus on creative work.
Devra is a versatile tool capable of handling software development and data files like CSVs and JSON. Unlike competitors like GitHub Copilot, Devra is platform-independent, supporting Mac, Windows, and Linux.
Please stay connected with AllAboutAI.com for more exclusive interviews and insights as we explore the latest advancements in AI-driven innovation!
