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.

Home Blog ...


Announcing 3 New Features to Accelerate the Software Development Life Cycle through Context and Codegen

Impactful code generation requires the injection of relevant context. These new features accelerate the SDLC with context and specificity.

1 July 2024

by Bryan Dunn

There’s no shortage of products being launched to improve developer productivity through the use of AI. At Crowdbotics, we’ve been intensely focused on making the SDLC more efficient through AI and found over and over again that the quality of AI outputs across all phases of the SDLC is dependent on great context.

Want to generate a PRD for an enterprise project? You’ll need to provide context about your specific infrastructure, services, vendors, security posture, and more.

Want to do impactful codegen and not parlor tricks? You’ll need to provide context about the entire project, including both technical and non-technical requirements. 

Our research shows that impactful code generation requires the injection of project-level requirements that AI can leverage to build what you want (not just line-by-line auto-complete). This includes context such as infrastructure, preferred languages, architecture, internal and external services, vendors and more.

Today, we’re happy to announce three big new features that accelerate the SDLC with context and specificity.

1. Enterprise Standards

First, we’re happy to announce that enterprise customers can now store and reference context relevant to their company, infrastructure, and applications via Enterprise Standards. This information is used by our CodeOps Platform to:

For example, let’s say you are a healthtech company that must maintain HIPAA compliance, is building on Azure using Node and Postgres, and has just procured Stripe to handle in-product payments for all your applications. Adding this information to your Enterprise Standards will now result in the following:

A PRD that includes Acceptance Criteria in line with HIPAA compliance
Technical recommendations of Azure-specific technologies that can handle HIPAA compliance

This context is referenceable by every generative AI feature inside Crowdbotics, ensuring that your team consistently builds to your standards. In the event that specific projects deviate from org-level context, your Enterprise Standards can also be modified for any individual project.

2. Technical Recommendations

The number one ask from users of PRD-AI has been to help fill the gap from PRD to execution. One specific area of feedback from enterprise customers is that suggesting specific services, vendors and technologies to fulfill the requirements would help shorten the planning phase before execution. For this reason, we’re happy to introduce the ability to generate Technical Recommendations.

In the screenshot below, you can generate recommended services to fulfill user account management along with links to product pages. Our goal is to make it easier to choose the right services and, just like modules, avoid building infrastructure and services that already exist. 

Technical Recommendations of Azure-specific technologies that can handle HIPAA compliance

3. Contextual Code Generation

Last, but not least, we’re closing the loop on the SDLC with Contextual Code Generation. Most code generation tools today start with the developer IDE and leverage the developer’s knowledge of the project to generate meaningful code. Our CodeOps Platform comes from the opposite end of the SDLC by building rich, precise requirements and helping generate plans for execution. What we’ve found is that impactful code generation requires all the context created upstream in the SDLC.

Our first release of Contextual Code Generation allows users to generate code for each Technical Recommendation. It leverages all of the context in your PRD and Enterprise Standards to produce impactful starter code in your desired languages and frameworks.

Copy and paste the code into your IDE to start using immediately. In the future, you’ll be able to push directly to GitHub and use the same context to make tools like GitHub Copilot contextually aware (and much more powerful).

These three features represent a major step toward comprehensive, accurate AI-powered app development for product and engineering teams. More to come!

Go from Idea to Code with AI