Web apps and AI products

A website tells someone about your business. An app does something for them.

We build the second thing: tools with a working brain behind the interface, shipped to real users. Ours don't stop at a demo in a deck.

What we build

Progressive web apps install to a home screen, work offline, and update without an app store review. For most products that's a better answer than native, because it ships in weeks instead of months and reaches everyone at once.

Internal tools and dashboards are the unglamorous software that decides whether your team has a good day. They're usually the highest-ROI thing a small business can commission.

AI-backed products cover retrieval over your own content, structured extraction, and assistants scoped tightly enough to actually be correct. We also build the unglamorous parts (evaluation, guardrails, fallbacks), because an AI feature that's wrong 5% of the time in public is a liability.

Content and publishing systems handle editorial workflows, membership and paywalls, and newsletter infrastructure.

Shipped

• Cuyar is a bilingual movement app for adults over 65. It's an installable PWA with 29 illustrated movements and offline voice coaching, and its design starts from the accessibility needs of the people who actually use it.
• Aguante is a bilingual fasting app built on one thesis: the popular claims in its category aren't supported by the evidence. Research is a feature.
• Pour AI is a wine assistant. Ask it a question or show it a photo of your dinner. It's scoped tightly enough to stay trustworthy in public.
• News4You is local journalism with membership, on a stack a two-person team can actually run.

How we think about AI features

Most AI product work fails in the same place: the demo works, and then real users arrive with real inputs.

So we build in this order. First we decide what "correct" means and how we'll measure it. Then we constrain the model to a scope it can actually be right about and ship a fallback for when it isn't, and only after that do we make it feel good. Anything else is a prototype wearing a product's clothes.

We'll also tell you when your problem doesn't need a model. A lot of "AI features" are a database query and a well-written rule set, and the version without the model is faster and cheaper. It's also correct every time.

PWA or native?

We build PWAs by default and tell you when a project genuinely needs native: deep hardware access, background processing, or App Store distribution as a business requirement.

What it costs

Web app and AI projects start around $15,000. The range is wide because "app" covers everything from a two-week internal tool to a nine-month product. We scope in phases, so you get something usable early instead of waiting on one big-bang launch. See pricing.

Common questions

Can you take over an app someone else started? Often. We read the code before quoting and tell you honestly whether continuing or restarting is cheaper. Sometimes it's restarting.

Do you do the backend too? Yes. Typically it's Supabase or Postgres with serverless functions. We set up row-level security properly, which is the step most rushed projects skip and later regret.

What about my users' data? We collect as little as we can and store it where you control it. Then we document it in a privacy policy that matches what the code actually does. We've shipped consent flows and deletion paths for products in regulated-adjacent categories.

Will it work offline? If it should. Cuyar works fully offline because its users are often in basements and community centres with bad signal. That was a product decision, and it came first.

Can it be bilingual? Yes, and it's cheaper to decide that on day one than to retrofit it. See bilingual work.