Jareth Purcell · Automation & AI systems engineer

I connect business systems and make AI useful inside them.

I design, build, and deploy production integrations for complex MSP operations. I turned an initial service-management MCP server into a reusable architecture for connecting AI assistants to business systems with the right access controls.

See selected work Get in touch

15+years across software, IT operations, and automation

MCPproduction integrations built from a reusable architecture

100+franchise scope of a separate solution I architected

Selected work

Systems built to do real work.

From secure AI access to billing data and cross-platform sync, I work through the details that determine whether a system holds up in production.

Production AI platform · deployed01

A reusable MCP architecture in production

I built a Python MCP integration for a service-management platform, then adapted its architecture for other MSP systems. The work spans API integration, user-aware access, deployment, and the operational fixes that make agent tools dependable.

Assistants→MCP pattern→Business systems

Deployed for operational use
Access tied to the individual user
Reporting tools that calculate results before returning them to an assistant

MSP data platform · shipped, evolving02

Automated billing data

Built a data pipeline that collects recurring-service usage from multiple vendors, brings it into a consistent model, and gives finance a clearer view of what should be billed.

Azure SQLRewstPower BI

Hands-on integration03

HubSpot ↔ ConnectWise sync

Built CRM-to-PSA synchronization for roughly 20 franchise organizations, connecting HubSpot and ConnectWise data across the two systems.

HubSpotConnectWiseAPIs

Solution architecture04

HubSpot ↔ Autotask

Served as solution architect for a separate bidirectional opportunity sync spanning more than 100 franchise organizations. Defined the data flow and rules for conflicting updates, and contributed technical work to the implementation.

ArchitectureData mappingPSA / CRM

Developer tooling05

HatchCheck

Built workflow linting and autocorrection tooling to catch structural problems early and make automation development more consistent.

Static analysisQuality systemsAutomation

Internal developer platform06

MCP server generator

Built a generator around the architecture proven across those deployments, giving new services a consistent starting point for integration and access control.

PythonMCPDeveloper tooling

Experience

From the code to the operating model.

I work across architecture and implementation: framing the operational problem, writing the integration, deploying it, and debugging what happens in production.

2026 — present

Automation & AI systems work · Forthright

Designed a reusable MCP architecture and deployed integrations built from it for AI access to MSP systems. Also built cross-system billing data workflows and cloud automation for operations.

2023 — 2026

Senior Automation Engineer · Rewst

Delivered production workflows joining identity, RMM, PSA, and CRM systems. Built reusable tooling, including HatchCheck, and guided solutions from scoping through support.

2017 — 2022

Software Engineer · CISO Global

Developed Python automations and reusable API libraries for licensing, billing, systems analysis, patching, and monitoring.

2014 — 2017

Senior Software Consultant · ConnectWise

Built custom applications and automation for major partners, combining SQL, platform expertise, and direct technical consultation.

Engineering approach

AI needs a dependable system around it.

Giving an agent a tool is the beginning. It also needs the right user permissions, bounded actions, reliable data, and a way to see what happened.

Connect

Connect PSA, RMM, CRM, billing, and cloud systems through APIs and well-defined data models.

Orchestrate

Move manual operations into repeatable workflows with clear state, ownership, and recovery paths.

Harden

Enforce user permissions, audit writes, monitor the data path, and plan for production failures.

Scale

Turn a working integration into a reusable platform and automate the next deployment.

About

Builder first.
Architect by practice.

I've spent more than fifteen years at the intersection of software, IT operations, and automation—from MSP systems and consulting to platform automation and cross-system architecture.

Today I build production tools that let AI agents work with MSP systems safely. I focus on the permission model, the quality of the data an assistant receives, and how the system behaves after deployment.

My background in cognitive science still shapes how I work. I care about the people inside the process, not just the diagram around it.

Experience acrossForthright · Rewst · CISO Global · ConnectWise · cloud systems

Technical focus

Tools are chosen to fit the system.

PythonPowerShellAzureAWSMicrosoft GraphREST APIsGraphQLSQLDockerGitAzure Container AppsRPAFastMCPPower BI