Crowdbotics Logo

Customers arrow

Don’t take our word for it, see what our customers have to say.

About Us arrow

We are on a mission to radically transform the software development lifecycle.

MATCHING CODE TO SPECS: MATCH-AI

MATCHING WITH THE CROWDBOTICS PLATFORM 

One of the pillars of the Crowdbotics platform and one of the keys to enabling a CodeOps practice is the ability to match requirements with  existing code. No matter how unique an application, there will invariably be some code that has already been written elsewhere. It might be as simple as an authentication dialog, or it might be for access to a particular database. For many applications, up to 70% of the code can be sourced from reusable code modules, but nearly every application can benefit from at least some existing code. 

There are four key elements that make up the mechanics of identifying reusable code: 

  • Included in the platform are libraries in which reusable code is available for reuse. 
  • The reusable code is stored in the libraries as ‘modules’.
  • The Match-AI function of the platform identifies modules that fulfill requirements. 
  • The requirements used by Match-AI come from the PRD that was generated by PRD-AI.

As a foundation of the matching process, it’s evident why improving the quality of PRDs is so pivotal to the success of a CodeOps practice. However, while high-quality PRDs are necessary, they are not sufficient by themselves. A centralized repository for the code as well as a mechanism to match requirements with code are also required. 

LIBRARIES 

There are two types of libraries: 

  1. There is a public Crowdbotics library accessible to any Crowdbotics platform user. The code modules in the public library come from Crowdbotics and our partners and have been validated by us. 
  2. There is a private library that is part of your private Crowdbotics platform and is only accessible by your staff. The code modules available in the private library are built, validated and managed by your team. 

These libraries are used by Match-AI to identify code modules that fulfill requirements in a PRD. 

The private library only contains the code module metadata. The actual code associated with the module resides in your current repository, Github for instance. The Crowdbotics private library contains information about the capabilities of the module, as well as other important metadata (e.g. author, date, version, etc.), and finally it includes a pointer to the location of the source code for that module.  

Finally the Crowdbotics platform includes tools that aid developers in building new modules and promoting them to the private library, as well as tools for an administrator to manage the library. 

On the next page, Assemble, we’ll address what is stored in each library – code modules.

MATCH-AI 

Like PRD-AI, Match-AI is made possible by the latest generation of AI. While simple in nature, Match-AI does an enormous amount of heavy lifting. It inspects every feature detail identified in the project’s PRD and then searches the public and private libraries to find the best code module to fulfill that requirement. If it finds a module, it will add it to the feature description in the PRD and it will include it in a consolidated project plan for the application. In concept, Match-AI is a rather simple step, but from a computer science perspective, it represents an enormous hurdle that few have been able to navigate. The introduction of the latest generation of AI models is what makes this finally possible.  

WANT TO GIVE MATCH-AI A TRY?

The Crowdbotics platform is available on a trial basis. Talk to one of our reps and we’ll get you access.