AI workflow automation & custom software for SMBs.
We help teams get practical value from AI, replace awkward manual processes, and build custom software where off-the-shelf tools fall short. When it ships, we keep running it for you, or your team can take it over: the code lives in your repository from day one either way. Start with a focused readiness session, a paid discovery sprint, a custom build, or ongoing care for software that already exists.
// for advisers Run a financial advice firm? There's a dedicated page for adviser portals →
AI readiness
For teams that know AI should be helping but haven't yet made it useful in daily work. We set Claude up for your team, connect it to the systems and documents you already use, train your people on where it's genuinely useful, and identify the workflows worth automating next.
- Claude set up for your team
- Connected to your existing systems
- Hands-on team training
- Shortlist of automation opportunities
Custom software builds
For teams with a workflow that shouldn't still be held together by spreadsheets, inboxes, PDFs, and manual re-entry. Every build starts with a short paid discovery sprint: we map the workflow, find the riskiest part, and turn it into a clear build plan before anyone commits to the full project. The recent benchmark is a financial-advice portal that replaced two paid SaaS products and reached paying-client use thirty days after it began.
- Discovery & written spec
- Weekly demos, your repo from day one
- Deploys & observability included
- 30 days post-launch support
Fractional CTO / AI adviser
For teams that need senior technical direction as well as hands-on delivery. We work alongside your team across AI strategy, build-versus-buy decisions, vendors, architecture, product scope, and delivery cadence, and still ship production features. Best fit when you need someone who can make the calls and still ship.
- Embedded on your stack
- Technology & AI strategy
- Specs, architecture & cadence
- Ships alongside your team
Ongoing care
Production software shouldn't be abandoned at handoff. For software we've built, or software we've agreed to inherit, we keep it hosted, monitored, patched, and improving: small changes shipped regularly and a clear path when something breaks. We're on call for your team and meet with you regularly to review what's working and plan what's next. It's how the portal client runs today.
- Hosting & monitoring
- Small changes shipped monthly
- On call, regular team check-ins
- Cancel anytime
A simple engagement model.
Discovery
A short paid sprint. We map the workflow, review the systems involved, find the highest-leverage target, and test the riskiest assumption before anyone commits to a full build.
Scope
Discovery becomes a written plan: what we are building, what is out of scope, the delivery stages, acceptance criteria, and timeline. You review it before the build starts.
Build
We build in visible stages with regular demos and async updates. Code lives in your repository from day one, with deployment and operating notes documented as we go.
Care or handoff
You get the code, the deployment process, the docs, and 30 days of support. After that, keep us on the care retainer to run and improve it, the way our current portal client does, or take it over yourself: everything you need is already in your repository.
Common questions.
- Why Bloody Good Software instead of a development agency?
- Agencies usually sell capacity. We sell a scoped outcome with the builder in the room. You deal directly with Cameron Carmody, the person doing the discovery, writing the spec, making the product decisions, and shipping the software. The recent proof point is a bespoke client portal for an Australian financial advisory firm that replaced two paid SaaS subscriptions and reached client use thirty days after the project began.
- Who is the right fit for an engagement?
- The best fit is a small or mid-market business with a workflow that has clearly outgrown the tools around it: onboarding, document handling, client intake, internal admin, compliance evidence, reporting, or manual re-entry between systems. We also work with consulting firms and in-house teams that need senior technical direction but not a full-time CTO.
- What does AI readiness include?
- It's the low-friction starting point for a team that wants to use AI properly. We set up Claude for Work and, where it fits, Claude Code. We connect it to the documents and systems you already run, train the team on practical use cases, and identify the workflows worth automating next. Depending on scope, this can be a one-day session or a two-week sprint.
- What kind of AI work do you actually do?
- Useful operational work: document extraction, pre-filling forms from existing paperwork, classification, inbox triage, workflow automation, internal tools, and AI-assisted review processes. We also build evaluation checks so we can test whether the system is reliable before it becomes part of a live workflow. If your idea depends on training a foundation model from scratch, we're probably not the right fit.
- How do you reduce key-person risk?
- Your code lives in your repository from day one, not ours. The deployment process is documented, the stack is mainstream, and the handoff materials are written as the project progresses. The goal is that another competent developer could pick the work up without needing everything explained from scratch. The ongoing care retainer keeps it maintained either way.
- Do you sign NDAs?
- Yes. We are happy to sign yours, or can provide a simple mutual NDA.
Got a workflow that should have been automated years ago?
Send through the workflow, the systems involved, and what currently breaks. We'll tell you quickly whether it looks like a readiness session, a discovery sprint, a build, or not a fit.