Enterprise data, automation and software · Miami, FL

Production systems for companies that have outgrown spreadsheets and point tools.

I design and build data platforms, SaaS integrations and custom software that run unattended, reconcile to the penny, and ship with tests. Fixed scope, fixed price, half up front.

4M+Florida permits indexed for PermitVista
50+public-records systems integrated
$0.00variance across 500+ reconciled fills
~85 msorder latency, live trading engine
10+ yrsPython, Azure, Salesforce, Snowflake

Services

Three ways to engage. Each one is a fixed-scope build with a deposit.

Data Platform Build

From $15,000

Pull your SaaS and operational data into one warehouse, model it, and put dashboards on top. Salesforce, HubSpot, call-center and LMS systems into Snowflake, Azure SQL or Postgres.

  • Incremental pipelines with change detection and retries
  • Nightly reconciliation against source systems
  • Dashboards and a documented data model
  • CI/CD, monitoring and runbooks

Fits: operations, finance and sales teams reporting out of exports and spreadsheets.

Automation & Integration

From $8,000

Connect the systems your team re-keys between by hand. Real-time webhooks, two-way sync, payment and messaging automation, with an event log so every action is traceable.

  • CRM, billing, telephony and e-signature integrations
  • Webhook event store and replay
  • Email and SMS workflows
  • Admin tooling and kill switches

Fits: firms with 3+ tools that don't talk to each other and staff who copy between them.

Custom SaaS & Platforms

From $25,000

A production application from scope to launch: multi-tenant backend, web app, payments, auth and deployment. Built the way I build my own products.

  • FastAPI or NestJS backend, React or Next.js front end
  • Payments (Stripe, Square, LawPay), SSO and roles
  • Automated test suite and release versioning
  • Azure or Render hosting with CI/CD

Fits: founders and operators with a validated idea who need a senior builder, not an agency.

Payment terms

Every engagement starts with a written scope and a deposit. Work begins when the deposit clears. Milestones are defined in the scope, so you always know what's next and what it costs.

50% to start25% at milestone25% on delivery

Process

How an engagement runs.

Scoping call

30 minutes. You describe the problem, I ask about systems, volumes and who touches the data. Free.

Written scope & price

Within 3 business days: deliverables, milestones, fixed price and timeline. No hourly surprises.

Deposit, then build

50% deposit starts the work. Weekly written updates, a shared repo, and a demo at each milestone.

Delivery & handoff

Tests passing, CI/CD in place, runbook written, your team trained. Optional retainer for support.

About

Jonathan Bush

Senior engineer and founder based in Miami Beach. I've spent a decade building and testing production systems for regulated and high-volume businesses: payments at Verifone, pharma at Bristol Myers Squibb, supply chain at Southern Glazer's, and healthcare CRM integration for a national veteran-services nonprofit.

Since 2020 I've run Jon Bush Enterprises and shipped my own products, so I build client systems the way I build things I have to operate myself: reconciled, monitored, tested, and documented. I'm currently CTO of Espiral USA and co-founder and builder of Moviz and PermitVista.com.

PythonFastAPITypeScriptNestJSNext.jsReactAzureSnowflakePostgreSQLAzure SQLSalesforceDockerKubernetesGitHub ActionsAzure DevOpsStripe
Company
Jon Bush Enterprises LLC, Florida, est. 2020
Current roles
CTO, Espiral USA · Founder, PermitVista.com
Education
B.S. Business & Computer Science, University of Tampa
Working style
Remote, US Eastern hours. Direct access to the engineer doing the work.
LinkedIn
linkedin.com/in/jbush-engineer

Contact

Tell me what's broken or slow. I'll tell you what it takes to fix it.

Scoping calls are free and take 30 minutes. You'll have a written scope and fixed price within three business days.

Book a 30-minute scoping call
Or message me on LinkedIn
Email: jon@jonbushenterprises.com
Based in Miami Beach, Florida · Working with clients across the US and Latin America
← All work

Case study · Proptech

PermitVista

Instant building-permit history for any Florida property.

Live · paying customersFounder & sole engineer
Visit permitvista.com
Scale
4M+ permits indexed · 50+ jurisdictions · results in seconds
Stack
Python · FastAPI · Azure SQL · Azure VM collector · Render · WeasyPrint · Square · SendGrid · GitHub Actions
Quality
CI pipeline · 71 automated tests

The problem

Open or expired permits derail closings. Agents, title companies, investors and homeowners used to check county portals one property at a time, each with its own search, its own formats and its own gaps. A single missed permit can hold up a sale or land on the new owner.

What I built

  • A FastAPI search engine merging a statewide permit index with 50+ county and city public-records integrations
  • Nightly bulk loads plus live lookups, with caching and de-duplication across sources
  • Automated open/expired flagging and branded PDF certificates ready for the closing file
  • Self-serve monetization: free teaser, per-permit or whole-address unlocks, and subscriptions, with Square payments and SendGrid email
  • CI with 71 automated tests, so every change to a jurisdiction integration is checked before it ships

Why it matters for your project

PermitVista is the same shape as most enterprise data work: dozens of inconsistent sources, a nightly bulk layer plus a live layer, de-duplication, and a product on top that people pay for. I built and operate every part of it.

← All work

Case study · Fintech

Espiral USA

A cross-border money app for US customers sending to Latin America, the UK and Europe.

Live · pre-launchCTO
Visit espiralusa.com
Scale
5 corridors · 100 PRs and 135 tickets shipped in 8 weeks · two-person team
Stack
TypeScript · NestJS · Next.js 16 · Prisma · Postgres · Azure Container Apps · GitHub Actions · PostHog · Bridge · Stripe

The problem

Immigrants and small businesses pay 5–8% and wait days to move money abroad. Espiral gives US customers an account, same-day sends, and a way to get paid in dollars.

What I built

  • NestJS API over stablecoin payment rails (Bridge) with a webhook event store, so every external event is recorded and replayable
  • Nightly reconciliation between internal ledgers and the rail provider
  • Per-corridor kill switches and admin tooling for operational control
  • Stripe card funding, Mexico receive via CLABE, and payroll groundwork
  • Bilingual Next.js app (Spanish and English), deployed on Azure Container Apps with GitHub Actions CI and PostHog analytics

Why it matters for your project

Money movement is unforgiving. Event stores, reconciliation and kill switches are what let a two-person team operate a payments product safely, and they are what I bring to any integration where the records have to agree.

← All work

Case study · Trading systems

Polymarket Edge

An automated market-making and research platform for Polymarket US, running unattended with real capital.

Live · productionSole engineer
Scale
58 markets quoted continuously · ~200 fills/day · ~85 ms order latency · 25,000+ snapshots per game
Stack
Python · SQLite · WebSockets · Azure DevOps · systemd · Deribit & Odds APIs
Reconciliation
500+ fills matched to exchange records, $0.00 variance

The problem

Prediction-market prices lag real information, and small traders can't win on speed. The fee structure rewards resting orders instead. Polymarket Edge earns the maker/taker gap with model-driven quotes and treats risk control as the product.

What I built

  • A 24/7 quoting engine pricing 58 Bitcoin and politics markets from Deribit options data
  • Live order-book stream client with order routing in about 85 ms
  • A fill ledger reconciled fill-by-fill against exchange records: 500+ fills, $0.00 variance
  • Circuit breakers for drawdown, exposure and stale feeds, with hard stops that survive restarts and deploys
  • In-game NBA and MLB market recorders with replay-based strategy testing
  • CI/CD to a cloud VM, deployed on every push and supervised by systemd

Why it matters for your project

This is the system I'd point to if you asked whether I can be trusted with production money. It runs around the clock, reconciles to the penny, and stops itself before a person has to.

← All work

Case study · Marketplace

Moviz

A peer-to-peer car rental marketplace for Guadalajara, with optional chauffeurs.

Live beta · soft launch nextCo-founder & CTO
Scale
~22,700 lines of code · 63 API endpoints · 129 automated tests · v2.8.0
Stack
Next.js · TypeScript · FastAPI · Azure SQL · Azure Blob · Entra External ID · Render · GitHub Actions

The problem

Guadalajara car owners have no local Turo-style marketplace, and renters need Spanish-first booking with deposits and optional drivers. Moviz is that marketplace, built for how rentals actually work in Mexico.

What I built

  • Listings and rule-based booking: hourly rentals, deposits, cancellations, late fees and odometer check-in
  • Travel zones with map pins; optional chauffeur and aircraft quote listings
  • Google and Apple sign-in on Microsoft Entra External ID; integrated payments
  • In-app chat with contact-detail blocking to keep transactions on-platform
  • Bilingual UI, CI, release versioning and a 129-test suite via GitHub Actions

Why it matters for your project

A two-sided marketplace is identity, payments, rules and trust all at once. Moviz shows the full build from auth to release process, shipped by a small team at startup pace.

← All work

Case study · Legal tech

JuryFlow

A multi-firm legal CRM, billing and client-intake platform.

In development · beta plannedCo-founder & engineer
Stack
Python · FastAPI · PostgreSQL · SQLAlchemy · React · TypeScript · Docker · pytest
Integrations
Clio · Filevine · Cerenade · LawPay · Affirm · Google Workspace

The problem

Law firms lose leads and fees across disconnected tools and manual follow-up. JuryFlow brings intake, case tracking and billing into one platform for firm owners, intake staff and billing teams.

What I'm building

  • Multi-tenant FastAPI backend with per-firm data isolation, JWT auth and role-based access
  • React dashboard: lead pipeline, case timelines and CFO analytics
  • Automated payment reminders by email and SMS; online payments via LawPay and Affirm
  • Two-way sync with Clio, Filevine and Cerenade; Google Workspace integration
  • Dockerized, with CI running pytest on every change

Why it matters for your project

Multi-tenant isolation and two-way sync with incumbent systems are the hard parts of any vertical SaaS. JuryFlow is that pattern applied to legal; the same architecture fits any industry with entrenched tools.