Back to the previous chapter: I - The specificities of building with AI
<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/d8baae53-7dc0-4bad-a65f-26898d6a633d/2df49ebd-c3ad-47c6-a91a-4ee5448a1e61/Silex_Brand_Symbol.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/d8baae53-7dc0-4bad-a65f-26898d6a633d/2df49ebd-c3ad-47c6-a91a-4ee5448a1e61/Silex_Brand_Symbol.png" width="40px" /> The best AI features are built as a team Bring together curious domain experts, product & design people and software engineers. Good news, you don’t need to hire “AI engineers” to start building.
</aside>
Roles and scopes will naturally vary depending on your company's resources, but it's important to recognize that beyond the exciting potential of this technology, what truly stands out is its accessibility.
Getting AI-generated output is as simple as calling an API. Any developer can do this—you don't need a team of machine learning engineers, data scientists, or data engineers to bring AI features to production. With a solid methodology and the ability to call an API, you're set.
As you tackle more complex use cases and depending on the complexity of your data stack, you may eventually need specialized expertise. However, remember that these experts are not a prerequisite to moving forward.
The prerequisite however is to acknowledge that building AI features is not only a developer job, a PM job, a data scientist job, it is a TEAM job.
→ Product
AI is a tool to solve customer pain points to generate value. The experts of user pain points are the product manager. They will ensure the AI feature is progressing toward generating value and a positive ROI for the company.
→ Design
AI powered features open a whole new field of questions for designer regarding user experience and interface. A chatbot may definitely not always be the right option to interact with AI, your feature UX and UI will probably require a design that keeps the “human-in-the-loop”, not for the sake of it but out of necessity for your system to be “grounded” by the user at certain steps avoiding loosing its way.
AI brings a new design paradigm and new patterns will emerge to fit the specificities of the technology. You need to involve design early in the process.
→ Engineering & Data science
