Head of Infrastructure
Take a running machine, understand every part of it, keep it healthy, extend it, and make it produce more with less human touch every quarter.
About the Role
AI Reputation helps clients control what shows up when someone searches their name, and increasingly what AI assistants like ChatGPT, Claude, Gemini, and Perplexity say about them. Search results are half the battle now. The other half is fought inside AI generated answers, and we are one of the few firms built for both.
We are not starting from zero. The backbone is already built and running: three automated prospecting engines that sweep DOJ, SEC, FINRA, state licensing boards, court filings, and press sources daily; a HubSpot CRM architected around our full sales motion; a client portal in active production; automated content publishing across multiple platforms and personas; and a Cloudflare Worker fleet handling call logging, notifications, and scheduled jobs.
What we need now is an operator to own it.
This is not a software engineering role. You will not be asked to architect distributed systems or write production application code from scratch. Our engineering work happens through AI development tooling, and you will use the same. What we need is someone who can take a running machine, understand every part of it, keep it healthy, extend it, and make it produce more with less human touch every quarter.
Practically, this role absorbs everything the CEO currently does outside of sales, under his oversight. Lead engines, client site builds, publishing automation, portal development priorities, integrations, fulfillment operations, and vendor relationships all move to you. That is a large surface area and it is the point of the job.
As the company scales, this role is expected to grow into the senior most technical seat in the company, including hiring and leading our first engineers. We call it Head of Infrastructure today because that is the work. Think of the trajectory as CTO.
The machine is already running.
Nothing on this page is a roadmap. The engines, the publishing system, the portal, and the Worker fleet run unattended today. You would be taking over a machine in motion, not building one from a deck.
The stack you would own, drawn honestly.
Select any part of the stack to see what it does and what owning it means. Every node here is in production.
Public sourcesDOJ, SEC, FINRA, licensing boards, press
DOJ press releases across all 94 districts, SEC and FINRA actions, state licensing boards, court filings, and press sources, swept every day.
Owning it means: sources drift. When a feed changes shape, you are the one who notices before the pipeline goes quiet.
Three prospecting enginesSweep, dedupe, enrich, route
Each engine sweeps a defined source set daily, deduplicates against the CRM before any spend, and enriches through a cost ordered vendor waterfall: free anchors first, paid vendors last.
Owning it means: daily operation, plus tuning the two numbers the business runs on: cost per qualified lead and reachable lead volume.
HubSpot CRMThe single source of truth
Architected around the full sales motion, it is the system of record for leads, clients, campaigns, and deliverables. Every other system reads from it or writes to it.
Owning it means: the schema, the properties, the routing, and enforcing how every integration touches it.
Outbound sequencingEmail, phone, SMS, LinkedIn
Qualified leads route to sequenced outreach with real deliverability hygiene: domain warmup, authentication, inbox rotation, bounce and reply handling.
Owning it means: deliverability, plus compliance with CAN SPAM, TCPA, A2P 10DLC registration, and platform terms of service.
Publishing automationScheduled AI agent tasks
Runs as scheduled AI agent tasks with per persona configuration, voice files, and topic calendars, publishing across Medium, WordPress, Blogger, and any platform we add.
Owning it means: reliability. Silent publishing failures are churn risk, so a missed post has to surface immediately, and the system has to grow without a linear increase in human effort.
Six publishing personasMultiple platforms, one voice each
Each persona carries its own voice specification and topic calendar, and the output has to read as a real practitioner writing, never as machine generated volume.
Owning it means: the quality controls that keep volume from ever costing credibility.
Client portalCloudflare Pages, D1, R2
Live and in active development: client authentication, credential vaults, image uploads, AI visibility scoring, and automated monthly scans.
Owning it means: the roadmap, the release cadence, uptime, data integrity, and the security of everything stored in it.
Cloudflare Worker fleetCall logging, notifications, cron
A fleet of Workers handles call logging, notifications, and scheduled jobs, all running unattended on a daily cadence.
Owning it means: webhook reliability, error handling, and alerting. You are the first responder when an unattended job does not run.
Zoom Phone integrationCalls become CRM records
An answered inbound call upserts the CRM contact within seconds, and completed call transcripts land on the record automatically.
Owning it means: keeping the handoff chain automated: booked call to Zoom link, call and recording to CRM record with transcript.
Drawn from the position description. The systems deep dive interview walks through the real thing.
What You'll Own
The Technical Backbone
- Own the existing stack end to end: Cloudflare (Pages, Workers, D1, R2, cron Workers), HubSpot, Google Workspace, Zoom Phone, Slack, Stripe, PandaDoc, and our enrichment and outreach vendor set.
- Maintain the data layer as the single source of truth for leads, clients, campaigns, and deliverables, and enforce how every system reads from and writes to it.
- Set and hold standards for security, access control, credential management, backups, and documentation.
- Make build versus buy calls and defend them on cost, speed, and maintainability. You own tooling and vendor decisions up to an agreed monthly threshold. Above that, bring a one paragraph recommendation to the CEO.
- Keep a running architecture document current enough that a new hire could orient from it in a day.
Client Site Creation and Asset Builds
- Own the production of client owned web properties: domains, DNS, hosting, SSL, site builds, and ongoing maintenance.
- Build the templating and automation that turns a client site build from a multi day manual project into a repeatable, largely automated process.
- Manage the full property portfolio across clients, including renewals, expirations, uptime, and technical SEO health.
Publishing and Content Automation
- Own our automated publishing infrastructure across Medium, WordPress, Blogger, LinkedIn, and any platform we add. Today this runs on scheduled AI agent tasks with per persona configuration, voice files, and topic calendars.
- Expand the system: new platforms, new personas, new clients, without a linear increase in human effort.
- Own the reliability of it. Silent publishing failures are churn risk. Build the alerting so a missed post surfaces immediately.
- Maintain content quality controls so volume never comes at the cost of the output reading as machine generated.
Fulfillment
- Deliver and oversee client reputation work across both search and AI surfaces: asset creation and publishing, profile and property builds, suppression campaigns, review generation, monitoring, removals and delisting requests, and client reporting.
- Build and run AI answer monitoring: track how clients appear in responses from ChatGPT, Claude, Gemini, and Perplexity, measure citation sources, and execute campaigns that improve how AI systems represent them (GEO and AEO).
- Turn fulfillment from artisanal to repeatable. Document the playbook, template the deliverables, then automate the parts that do not need a human.
- Own delivery SLAs and quality control. Build the dashboards that show campaign progress and client visible results without manual assembly.
- Manage and scale a small bench of contractors and virtual assistants against that playbook.
The Client Portal
- Own the product roadmap and release cadence for our client portal, currently live and in active development on Cloudflare Pages with D1 and R2.
- Existing surface includes client authentication, password reset flows, manager invite system, credential vaults, image uploads, AI visibility scoring, and automated monthly scans.
- Prioritize what ships next based on what reduces manual fulfillment work and what clients actually ask for.
- Own portal uptime, data integrity, and the security of everything stored in it.
The Automated Lead Engine
- Run and improve our three prospecting engines. Each sweeps a defined source set daily, deduplicates against the CRM before spend, enriches through a cost ordered vendor waterfall, and routes qualified leads to the correct outreach persona.
- Own the enrichment stack and its economics: free anchors first, paid vendors last, with a measurable cost per qualified lead.
- Own outbound sequencing across email, phone, SMS, and LinkedIn with real deliverability hygiene: domain warmup, authentication, inbox rotation, bounce and reply handling.
- Instrument the funnel so every stage from source to closed won is measurable, and tune against cost per qualified lead and reachable lead volume.
- Maintain compliance with CAN SPAM, TCPA, A2P 10DLC registration, and platform terms of service.
Systems Integration and Automation
- Keep the stack running as one workflow. Automate the handoffs: booked call to Zoom link, call and recording to CRM record with transcript, closed deal to onboarding and fulfillment kickoff, invoice to renewal.
- Own our payments and agreements chain, currently Stripe with ACH plus PandaDoc, and the automation that fires on deal close.
- Own partner and reseller integrations, including white label vendor APIs, data flow, and margin reconciliation.
- Build integrations through native connectors, direct API work, and Cloudflare Workers. We do not currently run Zapier, Make, or n8n, and we do not need you to. As we scale, adding a general purpose orchestration layer is on the table and will be your call to make and defend.
- Own webhook reliability, error handling, and alerting so silent failures do not become churn.
AI Agent Operations
- A meaningful share of our automation runs as scheduled AI agent tasks rather than traditional scripts. You will own that layer: writing and maintaining agent instruction files, managing the scheduled task roster, and debugging non deterministic failures.
- Know when an LLM beats a human or a rule, and just as importantly when it does not.
- Use AI development tooling as your primary build environment. Comfort here is not optional.
Security and Client Data Stewardship
- Clients trust us with sensitive personal information and, often, credentials to their own accounts and properties. You own the standards that protect it.
- Enforce credential management across internal systems and client accounts: shared vault, access tiers, onboarding and offboarding checklists.
- Define our data handling policy: what we collect, where it lives, how long we keep it, and who can touch it, with GDPR and CCPA awareness baked into how we scrape, enrich, and store personal data.
- Own backup, recovery, and incident response so a bad day is an inconvenience, not an existential event.
Coverage and Availability
Our lead engines, publishing tasks, and Worker jobs run unattended on a daily cadence. You are the first responder when they do not.
- Core hours are 8:00 AM to 5:00 PM Pacific, Monday through Friday, with real overlap with the CEO and COO.
- You own the on call posture for critical failures: what pages, how fast it is acknowledged, and where the runbook lives.
- Critical path failures, meaning anything that blocks lead flow, publishing, client deliverables, or payments, are acknowledged same business day.
How Success Is Measured
What We're Looking For
Required- 5 or more years across technical operations, solutions engineering, RevOps, agency operations, or a technical founder or early employee role.
- Hands on scripting and automation, Python preferred, and real comfort with REST APIs, webhooks, OAuth, and authentication flows.
- Demonstrated experience building with LLM APIs and AI development tooling. You should already be using AI to build, not evaluating whether to.
- Experience architecting and administering a CRM, not just using one. HubSpot experience is a strong plus.
- Built lead generation or data pipelines at volume, including scraping, enrichment, and validation.
- Comfort owning web properties end to end: domains, DNS, hosting, SSL, CMS, and basic technical SEO.
- Track record of shipping without a team behind you. You scope it, build it, and support it.
- Clear communicator who can explain a technical tradeoff to a non technical founder in one paragraph.
- Based in the United States and available during Pacific business hours.
- Background in ORM, SEO, digital PR, or agency fulfillment.
- Experience with GEO and AEO, LLM citation tracking, or AI visibility tools such as Ahrefs Brand Radar or Profound.
- Experience with content distribution, publishing platform APIs, search suppression, or review platform APIs.
- Cloudflare Workers, Pages, D1, or R2 experience.
- Experience hiring and managing offshore contractors.
- Basic front end skills for internal tools and client facing dashboards.
First 45 Days
This is an accelerated ramp by design. We expect full ownership by day 45, not a slow build to it.
| Timeframe | Outcome |
|---|---|
| Days 1 to 15 | Audit the existing stack, engines, publishing system, portal, and fulfillment process. Deliver a documented architecture map with priorities, costs, risks, and timeline. Credential vault and access standards verified and enforced. Full access handoff from the CEO complete. |
| Days 16 to 30 | Own the daily operation of all three lead engines and the publishing system outright. Alerting live on every critical job so no silent failure goes past one cycle. First measurable improvement to reachable lead volume or cost per qualified lead. |
| Days 31 to 45 | Fulfillment playbook documented with at least two workflows fully automated. Client site build process templated and largely automated. AI answer monitoring live for at least three clients. Portal roadmap defined with the first release shipped under your ownership. |
How We Hire
Short and practical. No algorithm puzzles.
- Intro call, 30 minutes. Mutual fit, your background, our roadmap.
- Systems deep dive, 60 to 90 minutes. Walk us through a system you built end to end: what it replaced, what it cost, what it saved, and what broke.
- Paid working session. A small, scoped, real problem from our stack. You are paid for your time whether or not we move forward.
- Offer.
Evaluate us the way we will evaluate you.
Technical candidates should judge a company by its systems. Here is what you would be walking into.
AI native from day one
Our engineering work happens through AI development tooling, and you will use the same. You should already be using AI to build, not evaluating whether to.
No legacy codebase
The backbone was built this year and runs daily: the engines, the publishing system, the portal, and the Worker fleet. You inherit a running machine, not an archaeology project.
A direct line, not a hierarchy
This role reports to the CEO and absorbs everything he currently does outside of sales, under his oversight. Decisions get made in one conversation.
The trajectory is CTO
As the company scales, this role grows into the senior most technical seat in the company, including hiring and leading our first engineers.
Shipped product you can look at right now: the client portal at portal.aireputation.com is live, in production, and in active development.
This is the job. Tell us why it is yours.
Resume, a short note on a system you built end to end, and any links you are proud of. That is the whole application.
Tell us what you built.
Send your resume plus a short note describing a system you built end to end: what it replaced, what it cost, and what it saved. Links to code, working automations, or live properties are welcome in place of a cover letter.
Application received
Thank you. Your resume and your answers are in, and a real person will reply to . Here is exactly what happens next, in order.