What I build
Four ways a service business runs better
Every service below is proven in a live, production system first — the case study is linked from each section. Texas-based, serving Texas and remote clients.
Custom CRM Builds
Burns Bright Studio builds custom CRMs for small and mid-sized service businesses — dispatch, billing, compliance, and sales pipelines designed around how the business actually operates, not a generic pipeline forced to fit.
- Built on production-grade stacks (FastAPI + PostgreSQL + React), not a no-code form builder
- Multi-tenant SaaS architecture when the CRM needs to serve more than one account (see RoofCRM: 13 live roofing-contractor accounts on one codebase)
- Real integrations — Stripe billing, Twilio, Clover, Samsara, RingCentral — wired into the workflow, not bolted on
- Incremental delivery: the Mac Septic CRM shipped 200+ API endpoints and 57 database migrations without downtime
Best fit: Field service, septic, roofing, manufacturing, and other operations-heavy businesses that have outgrown spreadsheets and generic CRMs.
Proof in production
TX Hail Leads
A multi-tenant SaaS platform that sells storm-damage leads to roofing contractors and gives each one a CRM to work them, built on top of government permit data.
Read the case study →AI Voice & Phone Agents
Burns Bright Studio builds AI phone and receptionist agents for service businesses — answering a real business line 24/7, not a chatbot demo.
- Built on Twilio → Deepgram → Claude → ElevenLabs (Pipecat), the same production voice stack running Phil, the agent answering Mac Septic's real line
- Model choice is evaluated, not assumed — Phil's model was selected via a harness scored against 51 real call transcripts (Haiku 4.5 beat Sonnet 98.1 to 95.8, ~30% faster, roughly a third of the cost)
- Handles appointment scheduling, after-hours intake, and status checks, with human hand-off when a call needs a person
- Cost: an AI receptionist replaces the cost of hiring a 24/7 answering service or an overnight dispatcher, typically at a fraction of that ongoing headcount cost — exact pricing depends on call volume and integration scope, discussed on a call
Best fit: Service businesses that miss calls after hours or during peak load, and can't justify a full-time answering service.
Proof in production
MAC Septic
Mac Septic's dispatch and sales run on a CRM I built from scratch, and their phone line is answered around the clock by Phil, an AI voice agent I evaluated into production against real call data.
Read the case study →Turnkey Digital Business
A turnkey digital business from Burns Bright Studio is a complete build for a founder starting from nothing: brand, an SEO-structured marketing site, a CRM, booking and payments, and the automation that keeps it all running.
- Brand and a cinematic, SEO-structured marketing site
- A purpose-built CRM for booking, client management, and day-to-day operations
- Stripe-powered booking with contracts and a client portal, live in production
- Nightly reconciliation automation keeping bookings, payments, and calendar in sync without manual reconciliation
- Proof: The Signature Pour, a Houston mobile-bar business that didn't exist until this build — brand, site, CRM, and payments, live
Best fit: A founder with an idea and no existing systems who needs the whole business built, not just one piece of it.
Proof in production
The Signature Pour
Brand, website, CRM, booking, payments, and the automation that keeps it running — a Houston mobile-bar startup that didn't exist as a business until this build.
Read the case study →AI Automation & Consulting
Burns Bright Studio is a small-business automation consultant that builds AI systems reading real documents — email, PDFs, insurance certificates — and turning them into structured, usable data, with production-data changes gated behind a dry run before anything writes live.
- AI extraction pipelines that read incoming email and PDF orders into structured production tracking (proof: a 35-year-old custom manufacturer doing $4.6M/yr, previously run off email and paper)
- Automated compliance chase engines — Whitedot's certificate-of-insurance escalation runs on a day 0/3/7/9 schedule with 10-point ACORD-25 document validation
- Statewide B2B lead scrapers and a 56M+ record, 50-state government permit-data pipeline feeding downstream lead products
- Every production-data change goes through a dry-run-first workflow — nothing writes to live data unreviewed
Best fit: Businesses buried in manual document processing or compliance chasing who need AI to do the reading and the chasing, safely.
Proof in production
Custom manufacturer (name withheld under NDA)
A custom manufacturer doing $4.6M a year still ran production off email and paper. I built the system that reads their inbox, structures the orders, and finds them new B2B leads.
Read the case study →Let's talk
Tell me about your business.
Fifteen minutes, no pitch deck. Tell me what's broken and I'll tell you honestly whether I can fix it.