ONE BIG READ

Dictate your work to your computer

Context matters. AI has demonstrated this for us again and again, like in this photo of “salmon” swimming upstream. 😆

In the past couple of weeks, we’ve seen startups trying to push AI further by making sure that it is context-aware. In our Product Hunt Dev tools newsletter, I wrote about Pieces Copilot+ and its Live Context feature, which captures context across everything devs are working on in real-time to help devs work faster and across tools.

Now TalkTastic is trying to tackle this for the rest of us with a context-aware AI voice keyboard app. It works by integrating across all your macOS apps, from Slack to Messages to your browser, to not only transcribe your speech with accuracy but refine and rewrite what you say based on your screen’s context.

So imagine you’re going about your day-to-day. With Talktastic, you’d hit the little microphone button and start dictating your response back to an investor in an email. The tool will write your reply while fixing any vocal tics, adjust words for tone, and spell the names of your investors correctly. Then hop over to Slack and shoot your coworker a DM and TalkTastic adjusts for that audience and context. In other words, TalkTastic interprets what you're saying based on what it sees on your computer screen.

TalkTastic “combines the capabilities of Apple Dictation, on-device Whisper, ChatGPT, Claude, and Google Gemini into one powerful, easy-to-use package,” explains the website. Founder and CEO Matt Mireles says the team plans to publish a scientific research paper on its “breakthrough in speech recognition.” It’s not Mireles’s first rodeo. He previously founded a Google Ventures-backed startup working on video transcription powered by speech recognition, which was acquired in 2012. He then went on to work as an investor and founded OASIS AI, which you can think of as a proof of concept for TalkTastic.

- Sarah Wright

PRODUCT HIGHLIGHT

Test and debug your APIs faster with this tool

If you're a seasoned developer, you know the importance of testing your code, especially if you're gearing up to launch a product to potentially millions of people. After years of building, testing, and releasing products to massive audiences, founders Abhishek Saikia and Sourabh Gawande decided to collaborate and explore ways to make the testing process more intuitive. The result is KushoAI.

Kusho is an AI agent explicitly designed for API testing. It helps developers automatically find bugs in their APIs before they deploy them. Drop in an API spec, and Kusho will generate an exhaustive set of tests based on real-world scenarios to identify any issues that could break your code in production.

From there, you can run each test individually if you want to go through everything with a fine-tooth comb or bundle them up and run them simultaneously. Kusho will then use AI to generate detailed assertions for each scenario so you can test the accuracy and reliability of your APIs.

Like other developer-orientated AI apps, Kusho uses a variety of services entwined together to create something bigger than the sum of its parts. Groq is one of them. It is an alternative to OpenAI, which Kusho founders described as a "blazingly fast large language model for code generation."

-Aaron O’Leary

AI IN THE HEADLINES

🤖 OpenAI defined 5 levels for AI to reach AGI.

🛍️ Amazon’s AI shopping assistant, Rufus, is now available to U.S. customers.

🛝Anthropic added a prompt playground to Claude to improve your apps.

👀 Microsoft has given up its observe role on OpenAI’s board, citing progress.

🍎 Apple has released the public beta of iOS18 but without any Apple Intelligence features.

📊 Microsoft has launched Spreadsheet LLM, a new AI built to handle spreadsheets.