SC Labs

Where ideas become experiments.

Not every idea is ready to be a product. SC Labs is where we validate, prototype, and sometimes abandon ideas before they consume real resources. This is the work before the work.

How the Lab Works

Three rules. No exceptions.

Validate before investing

We test assumptions with minimal code before committing engineering resources. An experiment should answer one question clearly, then stop.

Abandon without guilt

A failed experiment is a successful learning. We kill ideas that don't prove out. The goal is evidence, not sunken-cost protection.

Graduate with evidence

An experiment graduates to a product only when it has proven the hypothesis — not when it feels promising. Evidence over intuition.

Active Experiments

What we're testing right now.

Four experiments across four different hypotheses. Each one has a clear question it's trying to answer.

LegalRAG

Prototyping

Applying retrieval-augmented generation to U.S. personal injury case law. Can AI surface relevant precedents for case managers without replacing their judgment — or creating false confidence?

Python Vector DB LLM Integration

Testing with a curated dataset of PI case summaries. Evaluating relevance accuracy against attorney judgment on 50 test cases.

Atlas

Building

Internal tooling for managing SC Web Studio's own client project pipeline, invoicing, and deliverable tracking. A dogfood product — we build it for ourselves first, then decide if it's worth productizing.

Node.js SQLite Internal

Currently tracking 4 active projects through Atlas. Replacing the spreadsheet one row at a time. Hypothesis: replacing the spreadsheet reduces missed deadlines.

Topo

Exploring

A structured client onboarding system that captures project context, stakeholder information, and success criteria before any contract is signed. Replaces the discovery call document with something more systematic.

Forms Structured Data Automation

In exploration phase. Defining what structured onboarding actually needs to capture — and what it doesn't need to burden the client with.

Ledger

Incubating

Simple financial tracking for small Bolivian businesses that operate informally. Not accounting software — just enough structure to understand where money goes and whether the business is growing.

Mobile-First Spanish Bolivia

Early concept. Validating whether the problem is acute enough to build for. Conducting 15 informal interviews with small business owners in Cochabamba.

Graduated from the Lab

These experiments became products.

Not every experiment earns a product. These two did — because they answered their hypothesis with clear, positive evidence.

✓ Graduated · Now a Product

Case Assistant OS

Started as a single hypothesis: could a PI case manager's entire workflow fit in a local-first web app — no cloud, no subscription, no complexity? After 3 months of prototyping inside an active law firm, the answer was yes. It graduated to a product.

View Product

✓ Graduated · Now a Product

Project Home

Started as market research with one question: is Bolivia's rental market broken enough to warrant building a new platform? After 40 interviews with landlords and tenants across Cochabamba, the answer was unambiguous. Graduated to a product.

Join Waitlist

Follow Along

We document our experiments publicly.

Every significant finding, failure, and pivot is written up in Stories. If you want to see how a digital product studio actually thinks, that's where to look.

Read our Stories