AI content generation for API sales enablement

A latent gap in API sales

Enterprise companies with a global presence can struggle with growing turnover from API sales due to their sales teams’ focus and capacity.

It is a luxury and time-consuming to train a dedicated team of technical sales for the API program, and even more so to involve API experts and architects in every prospect’s pre-sales conversations. It is ineffective, and leads to miscommunication in the process:

The technical experts and architects might not focus on what sells to business stakeholders and why, while the business stakeholders are uninterested in raw technical information, and unrelatable use cases.

Additionally, developer portals have evolved not only to serve highly technical developers and architects, but also provide an understandable semantic layer for business personas, connecting the API and product capabilities to real-life scenarios, and business value.

The question of “How does the API work?” is complemented with “What does it mean for the business to use it?”. Without this layer, a business persona cannot clearly evaluate or discover whether a solution is offered to their specific needs.

Self-serve Business cases with the help of AI

Our solution combines specified and well-structured developer portal source content with sales context and AI content generation to create customized business cases that recommend products for specific customer pain points.

Source content

As a prerequisite, the feature needs a solid content architecture and information sources to provide a base for content generation:

• Content types and information architecture: The developer portal content is structured and written so technical specifications and business-focused content are present and distinguishable, while both being integral parts of the portal.
• Mitigate hallucinations: The content should serve as a sufficient source of information for AI-generation.
• AI-optimization and cross-linking: The content is optimized for LLM processing, containing sufficient context within sections, cross-linking related topics without having uncanny qualities for people who actually read it.
• Relationship mapping: API and product relationships and packages are not processed and interpreted by the LLM. The relationships can be mapped out in entity references in the content architecture. The LLM processes the input, and doesn’t infer connections itself.

Sales input

To further provide context and grounding for AI generation, our solution features a form where the account manager or salesperson can provide the customer’s context and input real issues.

This adds the relatable real-life use case that otherwise might be undocumented and missed by technical experts, while keeping the anonymity of the customer.

Content generation

The generation process takes place entirely within the portal, with all of its benefits:

• The system is prompted to propose the business value to stakeholders in your specific style and follows a predefined template for consistency.
• The feature is secure and protected using role-based access control, with complete control over data.
• The generated content can be edited, moderated, published, and exported in the portal CMS.