(Shipping) Out With The Code <> and In With The Doc šŸ“„

A dev's headache is a manager's must, so let's make that simpler, shall we?

For those that are new around here, this is weekly newsletter where I highlight new and innovative AI products that are worth exploring.

Hey hey!

Happy Friday! Weā€™re back for another Product of the Week!

Let Me Set The Stage After This Big Thank You!

Thanks for all your votes over these past weeks, with all your contributions, we found ourselves with a neck and neck finish for our 2nd week of ā€œPeopleā€™s Choice Reportā€. The poll said ā€œSam, we want dev tools featured, make it happenā€. And as a man of the people, I had to oblige šŸ˜‰

When it comes to Dev tools, there were a lot of different directions we couldā€™ve taken (according to my work, there were at least 17 subtopics in there I could work with), and I knew I didnā€™t want to focus on code generation. Respectfully, itā€™s been done, and done, and everything (and their dog) seems to be able to do it these days. Instead, I called in a few favors from friends to dig deeper into the reality of being a software developer. After a few greetings, my software engineer pals had plenty to sayā€”especially about code cleanliness. That got me thinking about how they handled their code reviews and why theyā€™re so painful. This train of thought eventually led me to a deeper issue at the heart of the developer experience: communicating a devā€™s intent clearly and explaining how the code works. And thatā€™s how I landed on the subtopic of the week: Saving developer time via documentation automation.

After testing a number of competitors, cue our Product of the Week:

Source: they themselves

Let Me Paint You Some Context

So. Like I was saying above, one of the biggest challenges developers face when bringing their work to the team and preparing it for production is making their intent clear on how they answered requirements ā€”communicating not just what their code does but why and how they did things. These same developers are paid to solve tough, complex, multi-part problems. But getting bogged down in tedious tasks, like writing documentation, is not only a waste of their time but also incredibly frustrating. Thatā€™s where AI steps in. The powerful capabilities of large language models (LLMs) are already known for their skill in code interpretation, and pushing them further to automate documentation can save developers significant time, allowing them to focus on more important tasks and boosting overall productivity.

No truer words have been spoken in a whileā€¦ Source: mimrr.com

Olaoluwa Ogundeji, the founder of Mimrr, said it himself that he envisioned a world where developers are empowered to focus on creative problem-solving, free from the tedium of manual tasks like documentation and bug fixing. (Source)

Solution Crafting

Mimrrā€™s approach is all about maximizing developer efficiency by automating the most time-consuming parts of their workflow. The paradigm of this solution is truly embodied by a developerā€™s response to being tasked with documentation:ā€Please, donā€™t make me do thisā€.

Needless to say Mimrr took that to heart. The core of the solutionā€™s design is a one-click , turn-key automated documentation system that generates a comprehensive repo document set without anyone lifting a finger. While the initial intake might take some time, the overall time savings are substantial, and the outputs generated are appreciable in accuracy and completeness!

The shape of this is setup to be a self-updating referenceable hub for in depth code.
What that means for the Admins is : Link itā†’ Set itā†’ Forget it. and for the Devs using it, it means: Read it, Review flaws with it, or Converse with it. (Yes, you read that right, youā€™ll see why below!)

Devs accessing Mimrr have 3 ways to get what they need. Whether it was intentional or not, the designers have included personal preferences in how information is absorbed!

Design & Usability

Mimrrā€™s user interface keeps things simple and intuitive, guiding admins through the setup process step by step. Onboarding is designed to be smooth, requiring just a few clicks to get started, and a relatively long coffee break to let the codebase get processed.

A very lean, purpose-driven admin panel for your appreciation

The interface doesnā€™t overwhelm users with options; instead, it allows them to engage with the tool as needed, making it easy to read documentation or dive into code insights at their own pace. This clean, streamlined approach ensures that 5 minutes into use, developers can get back on task, and start to flow with the tool.

What Makes It A Good Delivery?

This is a well buttoned solution that albeit some minor hiccups, does well in its functions. The tasks it handles are clear, outlined, and rightly dosed in their over-delivery to those expectations.

Thereā€™s no manual upkeep needed to have the docbase accessible to users, there are clear, simple means of accessing the outputs of the solution, and

using a built in chatbot to converse with the now generated code docs!

What sets this tool apart is the interactive feature, ā€œAsk Rune,ā€ which allows developers to engage directly with their codebase. Whether it's understanding how a feature was implemented or exploring nuances within large or unfamiliar codebases, this conversational capability opens up new possibilities. For newcomers or those dealing with complex projects, this functionality gives Mimrr an edge over other solutions, transforming codebases into more accessible and understandable environments.

Instead of devs spending hours manually documenting code, Mimrr delivers a starting point for every fileā€”minimizing the dreaded documentation task.

A point could be made that placing blind confidence in AI outputs could spell disaster, and to that I say: The business context may be in part missing, but the docs generated are almost guaranteed to be better than starting from scratch - or worse yet, putting a the codeā€™s author on Doc duty.

Letā€™s Get Into The Technicals

The Visible Tech Stack

This is a React.js Front-end, leveraging Microsoftā€™s ASP.NET framework, running on a Kestrel Web server.

Where AI Comes into Play

Need I say more? This involves a robust ingestion pipeline, directly tapping into an API integration with GitHub. The code interpretation happens, the code analysis for potential issues & dev best practices happens, and the data is made accessible to the chatbot. I suspect thereā€™s a branched processing tree of various LLM + RAG combinations for all of these tasks and then more.

Opportunities

Mimrr has a solid foundation, but its future growth looks even more exciting. Next up, it could make waves by integrating with more code repos, or then again, it could look at positioning itself as a more holistic Dev Documentation hub by entering the knowledge base tool market. If it syncs with Jira while bypassing Confluence, it might reshape how business and dev work connect. Capturing product leadershipā€™s decisions and bridging that gap between business context and development could redefine collaborationā€”potentially challenging heavyweights like Atlassian.

Closing notes

Mimrr is shaping up to be pain-relief for developers, cutting down on tedious tasks and boosting productivity. With its simple, and impactful automation and potential to bridge the gap between business and dev teams, it's one to watch as it evolves in the coming years!

TELL ME HOW YOU REALLY FEEL ABOUT THIS ONEā€¦ šŸ’Œ

Did YOU Like This Issue?

Login or Subscribe to participate in polls.

Have an idea to make this whole newsletter better but donā€™t want to be judged? Submit anonymously right here:

THE AI PRODUCT REPORT SERVICES!

Need a hand with your own SaaS? Letā€™s talk.

From Teardowns, and Business Fundamentals to Go-To-Market and, Product Growth, Iā€™m here to boost your business and help you crush your goals.

Check out what I offer and learn more here:

Letā€™s make something happen šŸ¤™šŸ½

Stay well, and until next week.

-āœŒšŸ½ Sam

Reply

or to participate.