@benthompsondev · Cambridge, Ontario · benthompsondev.github.io · LinkedIn

Healthcare IT systems, PowerShell and Python automation, DevOps growth, and practical software projects.

I work in healthcare information systems and systems support. The part of the job I enjoy most is taking disorganized operational work and turning it into something safer, repeatable, and easier for another person to trust.

Most of my strongest work is PowerShell automation around Microsoft 365, Active Directory / Entra ID, Exchange, SCCM, Intune, endpoint support, onboarding, migrations, reporting, and code review. I am building this GitHub as a public-safe version of that story: real workflow patterns, fake data, readable docs, validation checks, and projects that are useful enough to explain plainly.

I am aiming toward automation-heavy DevOps, cloud automation, platform/support automation, and practical software engineering work. Titles matter less to me than useful systems, good habits, accurate testing, and work that solves real problems.

⬇ Download SignalSpace Finance for Windows: private, local-first personal finance. Install one .exe, import a bank statement, and see where your money went. No account, no subscription, and your finances never leave your computer. See it first

▶ Try CloakScan in your browser — a local-first text redactor that cleans secrets and personal data out of text before you share it. Nothing to install, nothing leaves your machine.

I like building practical tools for the parts of IT work that are too messy to ignore and too repetitive to keep doing by hand:

• Disorganized CSVs that need validation before anyone trusts them
• repeated account, mailbox, access, and endpoint tasks
• migration projects where the same checks happen again and again
• reports and logs that project teams can actually use
• scripts that need guardrails before they should be run by someone else
• private text that needs a careful local review before it is shared
• documentation that makes a process repeatable instead of guarded knowledge

The pattern I keep coming back to is simple:

input -> validate -> plan -> review -> act or simulate -> log -> verify

That is the thread through most of my public work.

These are public-safe summaries of the kind of work behind the repos. Private workplace details, raw exports, internal names, and real data are intentionally not published.

This is my main portfolio repo.

It is built from sanitized versions of real workflow solutions I built in a large healthcare enterprise environment: onboarding, password remediation, Windows 11 device replacement, Chrome-to-Edge bookmark migration, workforce identity migration, O365 migration support, smaller troubleshooting utilities, and code review examples.

What I want this repo to prove:

• I can turn a manual process into a repeatable system.
• I care about validation, logs, state, rollback paths, and reviewable output.
• I can protect private data while still preserving the useful architecture.
• I can write docs and demo checks so another person can understand and run the work.
• I can review other people’s scripts and help make them safer before they hit production.

Good places to start:

• Script Index
• Reviewer Guide
• Public Release Checklist
• PowerShell demo checks

CloakScan is a local-first desktop app (Windows and Linux) for cleaning secrets, infrastructure details, and personal information out of code, logs, prompts, and support text before sharing them. There is also a live browser demo, so people can try it with nothing to install.

I built it because manually checking every script and log for hostnames, usernames, paths, tokens, and organization-specific details is slow and easy to get wrong. It runs without a backend or account and keeps scan content in memory only. The part I use most is the PowerShell-aware Portfolio-code mode: instead of turning a script into placeholder-filled soup, it swaps organization-specific terms for readable generic identifiers, so sanitized code is still worth reading — and v1.5 exports a small review kit (sanitized script, counts-only summary, manual checklist) to go with it.

This is currently my clearest end-to-end software project: React and TypeScript in the interface, a deliberately small Rust/Tauri desktop shell, a signed cross-platform updater, privacy-focused storage rules, detector and browser tests, GitHub Actions, release packaging, and an honest warning that automated detection still needs human review.

Good places to start:

• Try the live demo — runs in your browser, nothing is uploaded
• Download for Windows or Linux
• Screenshots
• Security and privacy model
• GitHub Actions

I built this for a real event, not as a sample storefront. It turns a wedding fundraiser into one complete flow for participants and the people running the draw.

Friends and family can choose entries, get the correct flexible price, submit a validated form, receive payment instructions, and see live aggregate totals. Payments happen separately by e-transfer and are checked manually. Behind the site, Google Apps Script records the submission in a private Google Sheet, sends confirmation emails, updates the public totals, and prepares one printable name slip for every included entry.

The final winner is drawn from the physical jar on video. The application handles the participant experience and the repetitive administration around the draw; it does not process payments or choose the winner.

Stack: React · TypeScript · Google Apps Script · Google Sheets · GitHub Actions · GitHub Pages

Good places to start:

• Live site
• Source code
• GitHub Actions

Private, local-first personal finance. A Windows app that turns your own bank and credit-card statements into a straight answer about where the money went, without handing your transaction history to anyone.

I built it because every budgeting tool I tried wanted an account, a subscription, and a copy of my finances on someone else's server. This one wants none of those. It is the project I use myself, most weeks.

Why you might want it

• No account and no sign-in. There is nothing to register for.
• No subscription and no service charge from me.
• Your finance database is a file on your computer, and the calculations run there too.
• Separate profiles keep two people in a household completely apart. Each profile is its own database, not a filter over a shared one.
• Backups, restore, and a full CSV export are built in, so your data is never trapped in the app.

Getting it running

1. Download SignalSpace Finance for Windows and run the installer.
2. If Windows shows a blue "Windows protected your PC" box, click More info, then Run anyway. (See below for why.)
3. Open SignalSpace. There is no sign-up, so it opens straight into the app and tells you what to do next:

4. Import a CSV your bank or credit card exports. Home and Insights fill in from there.

About that warning. SignalSpace has no Authenticode publisher certificate, which is the paid signing that stops SmartScreen appearing. Every release publishes a SHA-256 checksum you can verify the download against. Updates are a separate matter: in-app update downloads are cryptographically signed, and SignalSpace refuses to install one whose signature does not match its own key.

Bank formats vary a lot. SignalSpace reads common CSV exports and detects the date and amount conventions per file, but it deliberately refuses a file it cannot read confidently rather than guessing at your numbers. Some exports will need a manual column mapping, and some will not work at all.

What it does once your data is in

About the optional AI

AI is off by default and entirely optional. Every figure on every screen is calculated locally whether it is on or not. Turn it on and it explains findings SignalSpace already worked out. If you enable it, only the data you select for that question goes to the provider you configure, using your API key, and that provider may bill you for it. You can also point it at a model running on your own machine, in which case nothing leaves at all.

The only other network activity in the whole app is an update check that runs when you press the button in Settings. There is no telemetry and nothing about your finances is synced anywhere.

This is older and simpler, but it belongs here.

It comes from my Python PASS Leader / teaching support work. It shows where the coding side started: breaking down programming concepts for newer learners, writing beginner examples, and explaining code in a way people can follow.

My current work is in a large healthcare IT / information systems environment, mostly around Microsoft 365, hybrid identity, endpoint deployment, automation, documentation, and enterprise troubleshooting.

Before that, I worked across desktop support, networking, security software deployment, imaging, print services, licensing, and end-user support. I also worked as a Python PASS Leader at Mohawk College, where I created beginner programming material and helped students get comfortable with code.

Education and certification:

• Computer Systems Technician - Network Systems, Mohawk College
• Network Engineering and Security Analyst, Mohawk College
• Bachelor of Sports Business Management, Honours, Brock University
• CompTIA A+

I keep a Linux/networking homelab to learn by building real systems, breaking them safely, and improving the setup each time.

I want this GitHub to become a clean public trail of the kind of work I want more of:

• practical IT and systems automation
• cloud and DevOps-adjacent workflows
• useful PowerShell and Python projects
• local-first apps and tools
• CI checks, testing, docs, and privacy-safe examples
• projects that are easy for a teammate, recruiter, or anyone interested in tech to understand

The goal is not to collect the most repos. The goal is to show useful work that is organized, tested, and real enough to explain clearly.