aqua
Primary AI feature: Duplicate removal
More AI features: Test case generation, test steps, test description, test prioritisation, duplicate removal
aqua is an established Application Lifecycle Management solution. Launched over 10 years ago it is an all-in-one solution for creating products of any scale. It is a flagship solution of Cologne-based andagon GmBH that has served major Enterprises since 2001.
Benefits
- AI duplicate removal really adds up. Agile software development means that your requirements come and go, especially in a booming environment. A piece of tech that you considered two years ago might suddenly become more relevant again. Incomplete knowledge transfer may make a new Product Owner reinvent the wheel (and make it worse, too). Maintaining redundant code gets really expensive, too.
aqua’s AI can quickly sweep through your requirements and find ones that are a really close match. They will be highlighted on your screen to make a comparison and pick which one you would like to keep. If duplicate defects are anything to go by, you have about 20% redundant tickets as you’re reading the article.
- The AI suite goes beyond requirements. Testing features are the biggest highlight of aqua’s early AI launch. You can make entire test cases or fix them by just writing in plain English (or any language). You can populate tests with new test steps. You can prioritise tests to reflect your company’s individual QA experience. Client-tuned GPT-3 language model that you certainly saw in ChatGPT is magic.
- The pricing is friendlier than most solutions can offer you. Even if your team members need a separate licence to use aqua as requirements management software, a lot of colleagues won’t. Manual testers don’t need individual seats to run manual tests. People like designers can chime in on comments under requirements without using a paid licence. This is a welcome sight for SMBs, but also practically a unicorn in the enterprise world.
Disadvantages
- aqua is not a purely requirements management solution. While a major component of the ALM package, requirements-first tools like Jira can offer a better experience if managing features is your only goal.
- aqua is finishing a transition period when it comes to UI. A complete redesign is scheduled to finish by Q2 2023, and a couple of screens may look a bit jarring before that. UX, however, has already been polished in earlier iterations.
Get a testing strategy template that enables us to release 2 times faster
Notion
Primary AI feature: Requirements processing
More AI features: Copywriting
Planning-first platform, Notion has been adopted by development teams of various sizes as well. It is a relatively simple tool out of the box without a steep learning curve. The solution was launched in 2016 by an American startup.
Benefits
- Notion’s AI simplifies requirement creation. It doesn’t offer insight into software design, but definitely helps you organise and package your expertise. Here, requirements management with AI means summarising product meetings, highlighting action items, and even letting Notion write them out for you to edit them. There is a practical overlap with modern applications of AI in project management tools.
- Notion is a great knowledge retention platform. The mature wiki functionality enables your company to document work processes for new and current team members alike. This will also reap great benefits if Notion gets actually technical the way other AI tools for requirements management are.
- Notion is a competent mix of everything. While creative editing is still inferior to Google Docs, no issue management tool offers that many features for creatives. Software development planning can be better done in Jira, but not everyone needs to. You can even make a quick website if a market hypothesis warrants one.
Disadvantages
- Ticket hierarchy is Notion’s weak point. Even making subtasks was not an option until December 2022. The best you could do was mention subtasks within the primary task for quick navigation, and use labels to mark the primary task on the project overview. Trust our content team, we tried everything to make it work. It just didn’t.
The team behind Notion finally added subtasks, but they are far from straightforward. Each task can now be essentially labelled as a task or subtask, but it is one of the many properties that a Notion task can have. There is no snappy flow of adding, removing, and reallocating actual subtasks found in competitor tools.
- Notion won’t really help your coding. There is no native software development or quality assurance functionality. You could try to adopt some free or paid templates, but they are still limited. Running manual tests or making a bug report does not feel too different from forcing poor Excel to do the same.
WriteMyPrd
Primary AI feature: Requirements writing
More AI features: N/A
WriteMyPrd was released in January as a quick spin on GPT 3. The model behind ChatGPT was adapted to specifically tackle writing product requirements. The tool is co-developed by Olvy, the company behind an emerging user feedback processing tool.
Benefits
- WriteMyPrd aims to check all product requirement boxes. All output comes with a summary, goals, user stories, and individual requirements. It can also summarise scope, point out user expectations, and highlight dependencies.
- WriteMyPrd does solve the fear of the blank page for product owners. It can create some substance for making requirements out of plain non-technical description. Scope/out-of-scope pointers and user expectations can stir your creativity and help you cover more angles.
Disadvantages
- The output seems to lack actual depth. I have a buddy working on a pet project to make a better gaming-specific version of the reviews aggregator Metacritic. When I asked ChatGPT about it back in December, it came up with a list of features and technical to-dos. It also accommodated my buddy’s solution to review bombing and provided pointers to make it a reality.
On the other hand, WriteMyPrd did not tackle that solution well. It focused on a verified reviews experience from the crowdsourcers’ point of view only, neglecting how and why people would come to read such reviews. I then fed it a list of ideas from ChatGPT as features to see how WriteMyPrd develops them. The tool dwelled on stuff like using HTTPS and didn’t make an actual requirement.
For context, both ChatGPT and WriteMyPrd “understood” what Metacritic is. While WriteMyPrd offered a good structure, ChatGPT surprisingly provided more substance for actual product ownership.
- WriteMyPrd is not an all-in-one requirements management software. The tool expects you to save the output somewhere else. The only kind of export you can currently do is copy the text that you got and make the web page open Slack so you store the requirement there.
Conclusion
Requirements management with AI is a very new niche with few players. There are tools that offer interesting bits, but practically no one offers a complete AI-enhanced package. This will certainly be a potential market segment to follow.
Enjoy a proven AI-enhanced requirements management & ALM tool
On this page:
