nAtIve

Inspiration

Since AI showed up in our work lives, I've been watching my smartest colleagues hesitate to use it, especially the senior ones. It's almost always the same thing: either they write it off as a toy for the younger crowd, or they try it once, type in a question like it's a search engine, get back something generic that ignores everything they know, and decide AI isn't that great. Then they go back to what they were doing.

But here's what I kept thinking: a professional, for example a lawyer, already knows how to build an argument, weigh evidence, and anticipate the counterargument. So, I thought first, if experienced people could see AI in terms of what they already know, they could perhaps relate better and they'd be more curious about it. Second, if they could hand it a profile of how they think and want to work, and get some prompt profile they can reuse, they can have the outputs actually meet them at their level.

What it does

nAtIve is an AI fluency coaching web application that helps professionals become fluent in working with AI, starting from their existing expertise, not against it. It maps the user's existing professional skills and domain vocabulary to AI concepts, creating a personalized "translation map" that shows them they already think in ways that transfer to AI fluency. The product's tagline is: "You don't need to become someone else to be AI native. You just need a proper introduction."

It is mainly for working professionals in any domain (developers, lawyers, designers, analysts, etc.) who want to use AI more effectively but feel like a stranger to the technology, feel stuck, overwhelmed, or underserved by the AI output.

In a single no-account-required session, users get a personalized AI fluency profile including: a translation map connecting their domain expertise to AI concepts, an honest fluency assessment across 6 dimensions, an improved version of their own prompt with explanations in their vocabulary, and actionable next steps. Users can also export their profile as a portable "context card" to paste into any AI chat (ChatGPT, Claude, Gemini) so the AI understands their expertise from the first message. The key output is a portable text profile designed to be pasted into any AI conversation. This is the product's lasting value — it travels with the user beyond the session.

The product is grounded in Knowles' Andragogy (build on existing expertise), Sweller's Cognitive Load Theory (anchor AI concepts to existing schemas), Ryan & Deci's Self-Determination Theory (support autonomy, competence, relatedness), and Kolb's Experiential Learning Theory (connect experience to reflection to conceptualization).

How we built it

I don't have a technical background — I'm a legal professional who's interested in technology. So I built this entirely through vibe coding. I started by asking Claude to act as my engineer and brainstorm the idea with me. Once that conversation took shape, I refined it and used Abacus and Kiro to actually build the app around it. From there it was iteration many, many times over to clean up the code, deploying to Vercel, copy-pasting whatever errors came back, and asking the model to fix them. Basically the whole thing was prompting and iteration across different models until I had something that worked.

Challenges we ran into

Not being technical was the hard part. The AI could handle the actual coding, but the logistics around it were where I got stuck, like "deploying" to Vercel, "pushing" to Git, all the plumbing the model couldn't just do for me. Not knowing which files did what made it worse: I couldn't always tell what was safe to commit and what wasn't, so even simple decisions turned into guesswork and days spent to figuring it out.

Accomplishments that we're proud of

I'm proud that I actually deployed and shipped something myself with no coding background or any experience until I started vibe coding months ago! I am proud that I managed to figure Vercel, GitHub (push/pull/commit and all that confusing things!), and using an API key! I'm also proud of my idea for this product that is simple but impactful.

What we learned

Coding is not enough. The entire logistics around the product and making it deployed is not as easy as getting the agent help you with coding. So, I learned that you can maybe code something using vibe coding tools, but actually deploying a product is not yet completely "vibeable". Agents have not yet replaced the knowledge around product development and deployment logistics.

What's next for nAtIve

In my workplace I take a role as AI and Knowledge Manager. I'm thinking of introducing some of my colleagues to nAtIve and see if they can become AI native. I already had some users and they liked the idea. The output after using the simple context cards were considered as an improvement. So, I would like to work on this product based on more user feedback.

Built With

• anthropic
• css
• javascript
• nextjs
• tailwind
• typescript
• vercel