Launched this week
Fluent
Make your AI fluent in what matters
14 followers
Make your AI fluent in what matters
14 followers
Fluent works inside ChatGPT, Claude, and other AI apps, bringing in what matters for each question. A clothing purchase can draw on your closet and clothing budget. A meal plan can draw on your tastes, grocery needs, and grocery spending.
Hi Product Hunt, I'm Shane, the solo maker behind Fluent. I built it at night over the past six months because I kept asking AI apps for help with decisions that depended on more than the current conversation knew, such as what was already in my closet, what I planned to spend, how we like to eat, and what we needed from the store.
Fluent works inside ChatGPT, Claude, Codex, OpenClaw, and other AI apps that support MCP. It brings in the information that matters for each question. A clothing purchase can draw on your closet and clothing budget. A meal plan can draw on your tastes, grocery needs, and grocery spending. Each question uses the combination that makes sense for it.
I built Fluent three times. The first version was a full closet app for iOS and web. The second centered on its own agent. I scrapped both. Each version made the central idea clearer: the valuable part was a shared layer that worked inside the AI apps people already use.
One moment that made the idea click was a simple purchase question: should I buy another black tee? Fluent could bring my closet and remaining clothing budget into the same conversation. The assistant could see that I already owned several similar tees and that the purchase would use more than the budget I had left. That is the kind of answer Fluent is for: advice that can take more than one relevant thing into account.
Meals is the deepest area today. My assistant and I can plan a week, build the grocery list, and use preferences and grocery-spending context while deciding what to cook. When I later correct a recipe's timing or ingredients, that correction is available when the recipe comes up again.
Changes are shown before they are saved, and Fluent returns the saved result so the assistant can confirm what changed. You can inspect, correct, export, or delete the information you add.
The full build story is here:
https://meetfluent.app/case-study/
Try Fluent:
https://meetfluent.app
The Apache-2.0 runtime is available here:
https://github.com/shaner-git/fluent-oss
I'd especially value feedback on how quickly the product makes sense, whether setup feels straightforward, and which decision you would want Fluent to help with first.
genuinely interesting idea honestly, but i'd love to see a way to manually approve what context gets pulled in for each prompt. kind of worried about it grabbing the wrong budget data or whatever when i'm asking about something totally unrelated.
@nazlheol That’s a fair concern. Fluent doesn’t currently ask for approval before every read. Instead, your AI calls a specific Fluent tool for the task (Meals, Style or budgets), and Fluent returns the relevant context rather than everything you’ve saved. A closet question, for example, wouldn’t pull in your grocery budget.
A visible preview of exactly what Fluent shared is a thoughtful idea. I’m adding it to the launch feedback, thanks for raising it!
Would love to see a quick way to temporarily turn off certain data sources, like if I'm just brainstorming without wanting grocery spending influencing the suggestions. A simple toggle for each connected source would make it way easier to control context per conversation.
@bakar82283 That makes sense. Fluent doesn’t currently have a per-conversation toggle for each source. Context is scoped through the specific Fluent tool the AI calls, but temporarily excluding a source (especially while brainstorming) could give people more control. I’m adding this to the launch feedback. Thanks!
plugged it into chatgpt for a meal plan and it actually remembered my grocery budget from last week without me mentioning it, kind of wild how it just pulls the right context in
finally something that actually pulls in my real context without me copy-pasting it into every prompt. loved that it pulled my grocery budget when i asked for a meal plan.
@sefa1251873 Glad that worked well. Combining meal preferences, grocery needs, and budget context is exactly the kind of decision Fluent is meant to support. Thanks for trying it!
Really like how it pulls from my closet for outfit questions without me having to copy anything in. Saves me a lot of back and forth.
@azadkarasam Love hearing that. The goal is to make closet-aware advice available without making you restate your wardrobe every time. Thanks for trying Fluent!