WordPress DevOps: LiteSpeed, AWS SES, Cloudflare

ABOUT THE ROLE We are a small US-based international education company. We recently launched a new website with a HubSpot integration, so there is a block of work to get through over the next month or two. After that it settles into ongoing maintenance, occasional improvements, and being the person we can ask when something looks wrong. We are looking for a long term partner rather than a one-off fix, so please only apply if ongoing work suits you. Our website is a primary sales channel rather than a brochure site, so uptime and working forms matter to us more than they would for most small businesses. A few things to save everyone time: - Individual freelancers only. Agencies will be automatically disqualified. We want direct access to the person doing the work, and the person we speak to on the call should be the person doing it. - This is not a design or page-building role. If your focus is themes and layouts, you would likely find this work frustrating. - Most of what breaks on our site breaks below the CMS. We need someone comfortable on the command line who can tell a WordPress problem from a server problem. THE STACK - WordPress with the Divi builder, plus a child theme with custom code - OpenLiteSpeed on AWS EC2, with LiteSpeed Cache - Cloudflare in front (DNS, WAF, bot management, caching) - Amazon SES for transactional mail, via FluentSMTP - Amazon S3 for backups, via UpdraftPlus - Contact Form 7 for five public forms, with a custom server-side push into HubSpot - Weglot for multilingual - Wordfence for security - HubSpot as the CRM SAMPLE SCENARIOS These are the kinds of problems a qualified person should be prepared to handle and be able to reason about. If you read this list and have a sense of where you would start on each, we would like to hear from you. 1. A contact form reports success to the user but the notification email never arrives, and nothing is logged locally. 2. Recurring PHP worker exhaustion producing intermittent 503s, driven by automated traffic hitting URLs that do not exist. 3. A brief flash of unstyled content on page load, related to asynchronous CSS delivery and critical CSS generation. 4. Scrolling breaking on certain pages because of JavaScript event handlers bound at the document level. 5. Setting up SPF, DKIM and DMARC correctly, and moving a DMARC policy from monitoring to enforcement without breaking deliverability from Microsoft 365, Amazon SES or an ESP. 6. A custom CRM integration that only creates records, so a returning enquiry from a known email address errors out and is recorded nowhere. 7. Anti-spam honeypot fields that browser autofill populates, causing legitimate submissions to be rejected. 8. Configuring Cloudflare rate limiting and bot protection to keep automated traffic away from the origin server. REQUIRED We will ask about these specifically, so please only list what you have real experience with. - Linux server administration. Comfortable on the command line, reading logs, diagnosing resource exhaustion. - OpenLiteSpeed or LiteSpeed Enterprise. Worker and process configuration, LiteSpeed Cache behaviour, ESI, crawler settings, cache purging. This is the most important item on the list. - PHP at the code level. Not just installing plugins. You will write and debug hooks and filters, including Contact Form 7 hooks. - WordPress internals. The bootstrap sequence, the template hierarchy, must-use plugins, WP-CLI, the REST API, and what actually happens on a request to a missing URL. - AWS. EC2, SES (including production access, verified identities, bounce and complaint handling), S3, IAM. - Cloudflare. DNS, WAF custom rules, rate limiting rules, bot management, cache rules, page rules, and knowing when Cloudflare is the cause rather than the fix. - Email deliverability. SPF, DKIM, DMARC, alignment, and reading aggregate reports. You should be able to explain why SPF alignment is impossible for some ESPs and what that means for DMARC policy. - REST API integration work. Third-party APIs, error handling, retries, and logging failures somewhere a human will see them. - Backups and disaster recovery. Not just configuring them. Actually testing a restore. - English good enough for a live screen share where we are debugging something together in real time. NICE TO HAVE - Divi and the Divi builder specifically - Weglot or other WordPress multilingual setups - HubSpot API - Wordfence tuning - JavaScript debugging in browser dev tools - Basic frontend, HTML and CSS WORKING ARRANGEMENT - Ongoing retainer, not a one-off project. We want a long relationship with one individual. - You need to be available to start on the post-launch work in the coming weeks. - Time zone. We are on US Pacific time and generally available 6am to 8pm. We need a few hours of live overlap in your working day for troubleshooting together. Latin America and Europe both work well for this. Unfortunately Asian time zones do not. - We use BugHerd for tickets and email for everything else. - You will be the only technical person on this site, so you need to be comfortable owning decisions and telling us when we are wrong.