AI Prototype

Have an idea you want tested fast? We build Minimal Viable Product / Proof of Concept prototypes via a defined 2-week (+pre-planning) process. Turn your idea into a working minimal viable product in less time than it takes most agencies to schedule a kickoff.

How does it work?

1. 1 -
   We first schedule a quick, 30-minute introductory call to make sure the project fits the approach scope-wise, process-wise, and infrastructure-wise. Things that might make a project not work include too large of scope, not enough autonomy or decision makers in the process, or working off an existing product or technology that would need to be modified. These aren't all non-starters, these just need to be addressed ahead of time to make the project a success.
2. 2 -
   Next, we schedule a 3-4 hour meeting (online or in person) design sprint to get into the specifics of what will be built. We'll do a couple hours of work afterwards to develop Product specification & architecture. This meeting and work is billed at our normal hourly rate. Depending on the scope it may be for a scalable MVP (Minimum viable product) that you can take to market or more likely it may be a PoC (Proof of concept) that can be used to gather feedback and raise additional funds.
3. 3 -
   Finally, we carve out a two week block of dedicated time to work full-time on the project, with daily updates on progress. At the end of two weeks there will be a deployed product to review. The product spec will have outlined core requirements that will get built, and nice-to-haves that may get built depending on spec and churn.

The final output has a money back guarantee - if you're not happy with the quality or completeness of the outcome, you can walk away no questions asked. If you are happy with the output (and we suspect you will be) - we'll deploy it to your own AWS account, sign-over the code and GitHub repo and you can take the next steps in taking the product to market.

FAQ

How much does this all cost?

Cost will depend on a few factors - including the amount of Design work needed, but you can expect to pay around $25k all in. This cost is included in the product spec developed in Step 2 and won't change. We require payment upfront ahead of step 3.

What if I'm not satisfied with the output?

The contract we will both sign will stipulate you can request a refund after two weeks if you're not happy with the result. You'll just need to sign an agreement indicating you won't use any of the work that was built. We deploy to your AWS account & hand over the GitHub repos after the work is accepted.

What if the project isn't done in two weeks?

The product spec will lay out "requirements" vs. "nice to haves" - if the requirements aren't done at the end of the two weeks, we'll continue to work until they are. You can also pull the plug on the project and walk away with a refund. The "nice to haves" may not be done depending on how the project goes.

What kind of projects do you build?

Web & native Mobile apps are our bread and butter - we've been building them for over 20 years.

What tech stack do you use?

We focus on a few different architectures depending on the requirements and constraints. We deploy to AWS using a containerized or serverless approach so there's little to no IT burden. The most common stacks we use are one of Elixir/Phoenix/Postgres or Lambda/Go/Node/Dynamo. Frontends are built in React or SwiftUI (Transpiled to Android) These all scale extremely well and are AI-Development friendly.

Why should I pay you when I can vibe code it for free myself?

We highly recommend you do! Most UX is tough to reason about until you're actually using it. Vibing a prototype, or even just an artifact built by Claude can help you nail down exactly what you want to build and how users should interact with it. The more you figure out ahead of time the less we have to figure out as we're building and the more can get done.

The issue tends to be that building the core of a product without understanding anything of the software or architectural decisions leads to a brittle and non-scalable solution. Building on someone else's platform (another good vibing option) is great to get something out the door, but locks you into a specific ecosystem and vendor that may (and in this overhyped era very likely will) go away.

We've found the future-proof approach is to build applications using popular open-source languages, tools and frameworks and stay away from things that haven't stood the test of time.