According to TechCrunch, productivity startup Hero announced a new autocompletion SDK that automatically fills in AI prompts based on context. Founded by former Meta employees Brad Kowalk and Seung W. Lee, the company just secured $3 million in additional funding led by Forerunner Ventures after raising $4 million in seed funding last year. The SDK is currently invite-only and lets developers integrate prompt autocomplete into their apps, potentially making AI interactions up to 10 times faster by reducing back-and-forth. Hero engineer Saharsh Vedi developed the feature that predicts what users might type next in prompts. The startup is testing the technology in its own app for scheduling meetings and plans to release it to users within a couple of months. Hero is also in talks with ad tech startup Koah Labs to build AI-powered ads that appear in autocomplete suggestions.
Why this matters
Here’s the thing about AI prompts – most people are terrible at them. We type “book flight” and then stare blankly at the chatbot, wondering why it’s not reading our minds. Hero’s approach basically treats AI prompts like code completion in programming, where the system anticipates what you need next. This could actually make AI useful for normal people who don’t want to become prompt engineers.
And the timing is perfect. Every company is scrambling to add AI to their products, but the user experience often sucks. People try it once, get confused, and never come back. Autocomplete could be the bridge that makes AI feel less intimidating. Think about it – how many times have you abandoned a chatbot because you didn’t know what to ask next?
Bigger picture
This isn’t just about convenience – there’s real money at stake. Kowalk mentioned that companies operating at scale can save on server costs with fewer messages. That’s huge when you’re dealing with expensive GPU compute. But the real opportunity might be in those ad integrations they’re exploring with Koah Labs.
Imagine typing “book a flight to” and seeing airline suggestions pop up in the autocomplete. Or “find a restaurant that serves” with sponsored cuisine options. It’s either genius or slightly creepy, depending on your perspective. Either way, it creates a new advertising channel that’s deeply integrated into the user’s workflow.
Competitive landscape
Hero isn’t the only company thinking about this problem. Adobe’s Firefly app already lets users enter keywords in different prompt sections for creating soundtracks. And practically every major AI company is working on better prompt interfaces. But Hero’s approach of offering this as an SDK through their platform could give them an edge with developers.
The question is whether prompt autocomplete becomes a standard feature that big platforms build themselves, or if there’s room for a specialized provider. Given that Hero’s founders came from Meta’s AR team, they understand constrained interfaces better than most. That experience with limited screen real estate on AR glasses probably informed this whole approach.
Basically, we’re watching the evolution of human-AI interaction in real time. First we had command lines, then graphical interfaces, now we’re figuring out the prompt layer. And companies like Hero are betting that autocomplete will be a big part of that future.
