How to add Generative AI features to your SaaS product

Generative AI is all the rage these days. Every founder and product leader is asking themselves: “how can I add Generative AI to my offering?” And yet, only a few will get it right. Others will waste R&D effort and disrupt focus. 

So what can you do to make sure that your efforts are worth it? At LogicLoop, we’re some of the earliest adopters of LLMs. I’ve recently had the chance to present on a panel at Fintech Meetup on their implications for the fintech space. Here’s what I’d consider:

Understand your “why”

Before hopping on any trend, understand your Jobs to be Done. Your users are unlikely to directly ask you to integrate LLMs, so you’ll demonstrate proactiveness by improving their user experience and rate of content creation. Think about your business objectives – are they to improve user experience, improve funnel conversion or expand your audience? This will determine how you position, implement and price these features.

For instance, if you wanted to improve funnel conversion, you could offer generative AI features on your self-serve tier as well. But if you wanted to improve user experience at scale, you could reserve it for your enterprise tier.

Ensure you’re providing enough value

The #1 thing that startups integrating Generative AI need to remember is to add more value than being a thin layer on top of LLMs like GPT-4. Here’s some concrete ways you can provide value: 

1. Structuring the prompt: You can help your user get better results by providing a good user interface to structure their inputs, aka by helping them with prompt engineering. 
2. Integrate into their context: You can help your user productize their use of Generative AI by integrating into their datasets or workflows, so it’s seamless. 
3. Offer additional input and output: Go beyond what traditional models can provide with special visual themes, etc. that make your product stand out from competitors. 

Instrument, instrument, instrument 

With any new feature, comes the great responsibility of making sure it’s used and fulfills your business objectives. 

1. Understand usage: Log user activity to see what’s sticking and what needs more work. Supplement these quantitative metrics with qualitative interviews with users. 
2. Protect against bad actors: Make sure that you have logging and monitoring in place to flag any potentially harmful uses of your product. It can be too easy to do harm, unintentionally or intentionally. 

With that, enjoy the benefits of using Generative AI in your SaaS product! If you’re already integrated these features, can you share more on what’s worked?