Willow Voice
Voice dictation that actually works
Start new thread

Willow Knowledge - Your AI already knows you. Now Willow can too.

by
•
Willow Knowledge lets you bring your writing style and personal context into Willow from ChatGPT, Claude, Gemini, or any AI you use. It's like giving superpowers to Willow Scribe. From now on, your email, slack msg, prompt, or wherever you write will sound exactly like you when you dictate.

Add a comment

Replies

Best
Maker
📌
Long-time Product Hunt community. I'm back with another banger launch. Today, we’re introducing Willow Knowledge. Let's all admit: we talk to our AIs more than humans and they know more about us than we think :) They also know how we write and have context about most of our work. I talk to our users a lot and most of my calls end with this feature request: can Willow import context, my writing style from LLMs I use? We heard y'all and now Willow Knowledge lets you do that. For those new to Willow, Scribe our AI writing assistant, lets you speak the gist of what you want to say and write it for you inside the app you’re using. Knowledge gives Scribe extra special powers. It feeds more information about how you write and details about you that it can use when relevant. To get started, copy the prompt from Willow into the AI you already use, then paste the response back into Willow. It brings over your writing preferences and personal context. You can review everything, edit it, and add details yourself. > You can tell Willow to keep emails short, use your usual sign-off, or avoid certain phrases. > Your work msgs can have different preferences from your casual msgs. > You can also create app groups, the ones where you write other than email or slack. For ex: I've created Notion and Google Docs. > You can also add context about your work and company, so Scribe has that information when it writes a msg for you. Alright, so Willow Knowledge is available on macOS and Windows. Download Willow, import with the AI you already use, then Scribe emails, slack msgs like a PRO! Our team is hanging out here today and if y'all have any feedback or how we can improve Willow, plz share them in the comments below. Appreciate everyone checking it out.

 👋 Hey Sharath, good to see you on PH again. Cool launch too, make my AI writing more human!

Maker
appreciate you. Would love to hear your feedback 🙏🏼

Hi all, congrats on the launch! This is my first time learning about Willow. It sounds like your launch today is a plug-in or new feature to Willow called Willow Knowledge, where you can "export" (or copy & paste) your preferences from your everyday AI agent into Willow. Could I think of Willow Knowledge as agent skills for Willow?

Maker

Hi  thanks for checking out our launch. Knowledge is our new feature.

Yeah we can say that in a way. With knowledge import we are giving extra powers to Willow and when you dictate or use our AI assistant Scribe, it will write like you.

Maker

We’re launching Knowledge Base for Willow!

One of the biggest challenges with AI writing assistants is that they can write well, but they don’t always write like you. Over time, Knowledge Base helps Willow understand your writing style, preferences, and the facts and context that matter to you.

This is especially useful with Scribe, our AI writing assistant. You can tell Scribe what you want to say, and it can use that context to write something that sounds much more like you without needing as much editing.

Give it a try and let us know what you think. We’d love the feedback 🙌

I really like the separate preferences per platform. Because the way I want to write in Slack is different than in email, which is different than in WhatsApp.

One thing I'm finding frustrating with other AI tools is that it's not easy to switch between work and personal accounts, where I want to maintain separate context. Curious if you all are thinking about that. Some longstanding tools, like Linear and Superhuman, handle this well.

Maker

Hi  you are so right. We write differently in different places we work.

That's exactly why we built Scribe. It is our flagship writing assistant that reads the context on your screen on top of the instructions you give and uses knowledge from the LLMs you use to write exactly like you :)

You should give it a try. Not glazing but it's pretty magical.

banger launch 🫡

Maker

A friction-removal product with a friction-maximal front door

I had to fight to find the download, then fight through 10+ onboarding screens before I could touch the actual product. For a dictation tool. Let that sink in.

The fight to find it: no pricing anywhere on the site, a "Download for Free" button that redirects to this Product Hunt page instead of downloading, and an email-gated download that stalled — I ended up on a third-party mirror.

The fight to try it: screen after screen of onboarding theater, ending at "2 ways to use Willow" where the Dictation card is dead on click. The only way through was the Scribe path. Of course — Scribe is the metered feature (20 uses/week on free) and dictation is the unlimited one. The working button is the one that rations you.

I dictate all day and came here rooting for a native Windows alternative to Wispr Flow's wheezing Electron build. I never got to judge the transcription, because the company wouldn't let me reach it. A friction-removal product with a friction-maximal front door. I'll try again when the door opens.

I’ve actually used Willow Scribe before and one issue I noticed was that sometimes the context I had in mind was different from what it picked up

So Willow Knowledge feels like a really useful upgrade

Curious how it handles personal context across different AI tools

Really excited to try this one and congrats on the launch!

Maker
thanks. Yes plz give it a try. It’s a whole new way of dictating :)
Maker

 Thanks Suryansh! Would love to hear how it works for you!

bringing writing style in from ChatGPT/Claude/Gemini is a smart move, that's context people already built up instead of starting from zero. one thing I'm curious about: I write differently in a Slack message than in a client email, does Willow pick up on that and adjust per context, or does it blend everything into one averaged "you" voice? that distinction mattered a lot when we built something similar for customer-facing text at Dial, a single blended voice ended up sounding a bit off in the more formal contexts

Maker

 yes we designed Willow in a way that has context of your screen, which app you are using, and generate output in a way that 100% you

 good, screen context as the signal is simpler than I expected, didn't realize it was reading the app itself rather than you having to tell it which mode you're in. that's probably the right call, since asking the user to manually switch "tone" before dictating is exactly the kind of extra step that kills adoption for something that's supposed to feel invisible