Working for a US company with several, very different, lines of business taught me how difficult it is to post on social media consistently and keeping the essence of each brand in all posts. Standard AI tools helped me get the first draft, but that was it. I would spend hours editing the output to make it sound a bit more like me, but in the end this was more time-consuming than starting from a blank page. Tired of choosing between consistency and maintaining the brand’s essence and my voice, I built Riley. My goal was to teach AI how I sound and the essence of each brand to remove endless prompts and constant tweaking, keep creative momentum, and make content creation a smooth and enjoyable process. Riley learns from previous content you’ve written to generate pieces that sounds like you, not like AI. You can import examples (LinkedIn posts, website copy, social media, etc.), define context with @context-pills for background, audience, and style, and refine with feedback and comments. With one-line prompts, Riley outputs tailored pieces for multiple platforms at once while adapting over time. The more you teach Riley with feedback, the better she’ll write. Riley learns your style by extracting patterns from your imported writing and ongoing edits/comments, such as vocabulary, sentence rhythm, tone preferences, structures, and what to avoid. No need to actually explain to her “how you write,” she will know the more you work with her. This becomes persistent context that informs every generation, rather than one-off instructions. How is Riley different from ChatGPT or Claude? Standard LLMs require repeating lengthy context in every new session, and if you ask the model to create several pieces at once, the LLM will most likely write 2-3 really good pieces, and then start creating content you wish the LLM had never written. Claude and ChatGPT (and all other LLMs) don’t retain or iteratively improve from your feedback across conversations in the same seamless, workspace-based way. Riley keeps your writing DNA and style active and refines it continuously. Looking for feedback about (1) onboarding flow, (2) platform UX/UI, (3) how well Riley preserves nuances of personal/brand voice, (4) features you’d like to see in Riley, (5) anything else you can think of. Thank you HN community!