Learning to use AI and build software
Practice on real tasks with AI assistants, coding agents, testing, Git and deployment. Review the result and learn to maintain it.
See how it works/services / Vertex Authority
AI for Business
Integrations for customer inquiries, lead intake, document search and repetitive administrative work. Start with existing systems and a defined result to test.
Training covers independent deployment and maintenance of AI systems. Full-service setup, testing and ongoing support are also available.
Services
AI training, integration and custom development.
Assess a workflow, connect existing systems and choose between independent deployment training or managed implementation.
Figuring out where to begin
Review repeatable tasks, existing software and data. Identify automation candidates, human review requirements and a first pilot.
See how it worksWebsite chatbots and agent integrations
Handle customer questions, collect inquiry details and find approved information. Connect forms, email, calendars, CRM and internal tools, with permissions, checks and a way to hand work back to a person.
See how it worksCustom software and websites
Build an internal tool or website when existing software falls short. Work in client-owned repositories and accounts with setup notes and optional code training.
See how it worksWorking together
One workflow, a working example, then everyday use.
Agree on a scope, test a working example and document deployment and maintenance.
- 01
Talk through the work
Discuss a repeated task, the people doing it and the tools involved. Agree on what a correct result looks like, which actions need review and the setup and running costs.
- 02
Build and check a small example
Build a first version using approved examples. Test successful cases, failures and handoffs to a person. Include code and deployment training when needed.
- 03
Put it to use and keep it working
Connect the workflow, document the setup, and review errors and usage. Assign ownership of updates and support, with training or managed maintenance.
/services / setup & support
From a first example
to day-to-day use.
A working demo is a starting point. Test cases need to check whether the task was actually completed, what happens when an answer is wrong, and how a person takes over. Important actions use agreed review steps and permissions.
Keep the first project small enough to test and budget for. Compare setup, software, model usage, hosting and support costs before expanding it. Reuse existing tools where they fit and build custom pieces where the workflow needs them.
Keep code in client-owned repositories and accounts, with setup notes and an agreed support plan. Training can cover code changes, coding agents and deployment, or maintenance can be handled as a service. Websites, internal tools and private AI are also in scope.
/development / in progress
Website chatbots
& reusable agent tools.
Development is underway on Claude-powered website chatbots and reusable agent integrations. The aim is to adapt common pieces to a business’s information and workflows, with less setup each time. These systems are still in development; a self-service product is not publicly available yet.
The work includes conversations, document retrieval, tool use and visual interaction when an app has no suitable API. Testing focuses on task completion, errors, costs and review before consequential actions. It includes both open-source utilities and proprietary integration code.
MacShot and SideScreen remain public MIT projects. Their download and source pages are available independently of business services.
Software & source →/services / time & cost
Some ways we can work together
Current starting ranges are below. We agree on the work and cost before beginning. Third-party software, model usage and hosting are separate. Custom teaching or software work can be scoped around your team.
Introductory conversation
A free 15-minute conversation about what you want to learn or build, and what help is needed.
No charge
AI Essentials Workshop
Hands-on practice with AI tools, coding agents and examples from your team’s work, including how to check results and protect data.
Starting range: $750 to $1,250
AI and Workflow Assessment
Look at how the team works, which tools and data are involved, and a sensible first example to build.
Starting range: $1,250 to $2,500
One Workflow Pilot
Build and test one controlled system, such as lead response, intake, document extraction, knowledge search, or reporting.
Starting range: $2,500 to $5,000
Business AI Integration
Expand a proven pilot across the systems, teams, permissions, reporting, and operating procedures needed for daily use.
Starts around $5,000
Website and AI Integration
A custom website or redesign connected to forms, booking, CRM, payments, analytics, support, and follow-up.
Starting range: $3,750 to $12,500
Managed AI Operations
Help maintaining the software: tests, bug fixes, usage costs, provider changes, documentation and time for team questions.
Starting range: $400 to $1,500 per month
Questions
Questions about working together
Can you help our team build its own software?
Yes. We can work through a small app together, including requirements, coding agents, code review, testing and deployment. Sessions can start at the staff's current skill level, with setup notes and time to practice changing the result.
Do we need AI at all?
Not always. A normal automation, a better form, a software setting, or a clearer process may solve the problem more reliably. The assessment is designed to rule out bad AI projects as well as find good ones.
Do you train employees who are completely new to AI?
Yes. Training can start with a practical overview and then move into role-specific examples, safe-data rules, verification, approved templates, and follow-up office hours. The goal is useful habits, not a one-time prompt lecture.
Which provider or model do you recommend?
The answer depends on the workflow, current Microsoft or Google environment, number of users, data sensitivity, integrations, quality, latency, and budget. Evaluation covers business plans, model gateways, open models and local runtimes. Model availability and pricing change quickly, so they are verified again when the project is scoped.
Can you build a website too?
Yes. Websites are a major part of Vertex Authority. They often become the customer-facing layer for lead capture, booking, payments, chat, portals, analytics, and automated follow-up.
Can the system run locally or privately?
Sometimes. Local or private deployment can be appropriate for sensitive documents, predictable workloads, offline use, or vendor-control requirements. It also creates hardware, patching, access-control, backup, logging, and support responsibilities. The requirements need to justify those tradeoffs.
Will AI be allowed to take actions on its own?
Only within an agreed boundary. High-impact actions should use permissions, approval steps, logs, validation, and a clear way to escalate to a person. The goal is useful automation, not unchecked autonomy.
How is the work priced?
Current starting ranges are listed on the Services page. Workshops and assessments are fixed scope. Implementations are usually fixed price or milestone based. Ongoing monitoring and improvement are monthly. Software, model usage, phone minutes, and other third-party costs are listed separately.
/services / a first conversation
Project inquiries
Send a short summary of the workflow, training question or software project. A finished brief is not required.
Replies normally arrive within one business day. A free 15-minute call is available. Scope and cost are agreed before paid work begins.
Request a call →Project fit check
Direct with ZachA question is enough to start.
Send details of what you want to learn or build, or request a free 15-minute conversation.
15 minutes · Free · Video or phone