Claorova

/

Live work, shipped.

Live, launched websites, brands, and Shopify stores we built for barbershops, towing, lawn care, tax and accounting firms, electrical contractors, cleaning companies, concierge brands, fitness coaches, and e-commerce labels across Phoenix, Scottsdale, Tucson, Miami, Tampa, New York, and Orange County.

A selection of what we can show publicly. These are samples of our work, not all of it; much of what we build stays confidential or under NDA.

/tools · engines and systems

Tools we built.

Beyond websites, we engineer the products and internal systems behind them: estimating engines, automation pipelines, finance platforms, and simulation tools.

AI estimating engine

The Volt Planner

We built an app that turns an electrician's job details and floor plans into an accurate, code-checked price quote in minutes, instead of the hours it usually takes to work out by hand.

Rough estimate: bids that took hours now take minutes

Job details and floor plans into code-aware bids in minutes.

An AI estimating and permitting platform for electrical contractors. Purpose-built residential and commercial engines turn job details and uploaded floor plans into accurate, NEC-validated bids, with jurisdiction-aware permit packages and a companion mobile app for the field.

  • Residential and commercial estimating engines with material takeoffs and labor curves
  • Floor-plan and takeoff extraction from uploaded PDFs
  • NEC code validation, automated fixes, and electrical load calculations
  • Jurisdiction-aware permit packages and AHJ submission profiles
  • DXF drawings, branded PDF estimates, and QuickBooks export
  • Companion iOS and Android app with offline sync and on-site sign-off
Next.jsTypeScriptAnthropic SDKPostgres + DrizzleStripeCapacitor
Visit the platform
plan
estimate
NEC
permit

Automated tax-return engine

MasterTax Engine

We built a tool for an accounting firm that reads tax documents like W-2s and 1099s and does the math for a full federal, New York State, and NYC return automatically, with a person checking every number before anything is filed.

Rough estimate: most of the manual prep time per return, gone

Source documents into a computed Federal, NYS, and NYC return.

An internal tax-preparation engine for an established US accounting firm. A vision-AI layer extracts data from W-2s, 1099s, and K-1s, while a separate deterministic engine applies every bracket, deduction, and credit as auditable law-as-code, so no AI-generated number ever lands on a return line. Every figure carries a provenance trail and a mandatory staff review before filing.

  • Vision-AI extraction of W-2, 1099, K-1, and other source documents into typed fields
  • Deterministic Form 1040 engine: Schedules A/B/C/D/E/SE, QBI, AMT, NIIT, SE tax
  • New York State (IT-201/IT-203) and NYC computation, PTET, and Yonkers
  • Depreciation and business-property modeling (MACRS, Section 179, Form 4797)
  • Per-line provenance trail with mandatory review-before-file workflow
  • Build gate requiring every tax constant to cite an IRS, NYS, or NYC authority
Next.jsTypeScriptCloudflare D1 + R2Vision LLMZod
AI vision · extract
deterministic · rules
1040 · A/B/C
NYS IT-201
NYC · PTET
return

Internal finance platform

QNOW Ops

We built one place for a consulting firm to bill its clients and keep its books, replacing a stack of spreadsheets and a couple of separate apps.

Rough estimate: hours of monthly admin saved

One place to bill clients and keep the books two ways.

An internal operations platform for a New York City financial consulting and fractional CFO firm. It pairs an invoice and client manager with a dual-mode bookkeeping system, so the practice tracks engagements, bills clients, and keeps books in both cash and accrual views from a single tool.

  • Invoice and client manager with engagement tracking
  • Dual-mode bookkeeping: cash and accrual views from one ledger
  • Billing, payment status, and client records in one place
  • Reporting built for a fractional CFO workflow
Next.jsTypeScriptPostgreSQL
invoice#0421
paid
cashaccrual
net$ 84,200

Optimization & simulation engine

Yield-Uncertainty Supply Chain Simulator

We built a simulator that stress-tests computer-chip supply chains to find where they would break under real-world uncertainty, so planners can prepare before it actually happens.

Rough estimate: thousands of what-if scenarios modeled in minutes

Stress-testing chip supply chains against yield uncertainty.

A numerical simulation engine that models a three-stage semiconductor supply chain, from fabrication to die bank to assembly, operating under uncertain manufacturing yields. It runs hundreds of controlled optimization experiments that sweep uncertainty levels, pinpoints the tipping points where ignoring uncertainty starts costing real profit and service, and distills the results into a clear decision rule.

  • Three-stage supply chain modeling, fabrication through assembly
  • Stochastic vs. deterministic planning comparison
  • Automated parameter sweeps across hundreds of scenarios
  • Failure-threshold and cliff-effect detection
  • Cost-structure robustness and sensitivity analysis
  • Decision-map and lookup-table generation from results
PythonPuLPCBC solverNumPyMatplotlib
FABDIE BANKASM
profit · servicecritical thresholdyield σ →