AI Agent Developer & Operations Specialist

Do NOT send a generic cover letter. We will ignore it.
Step 1: Fill out the application form here: ---------- — attach your resume/CV.
Step 2: Record a 5–8 minute video answering the four questions below. Upload it to Google Drive, Loom, or YouTube (unlisted) and paste the link in the form.
VIDEO QUESTIONS (all four are required):
• Show us your screen and walk through a project where you built or deployed an AI agent or automated workflow. What did it do, what tools did you use, and what broke along the way? We want to see a real project, not a slide deck.
• Our managers are not developers. Explain what an API is and why it matters — as if you are talking to someone who has never written a line of code. This tests whether you can communicate with our team.
• You are tasked with building an AI agent that automatically sends a follow-up email to a client 3 days after a deal moves to "Closed Won" in a CRM. Walk us through how you would build that — what tools, what steps, what could go wrong, and how you would monitor it.
• What AI tools or frameworks have you used in the last 90 days? Show us something you are currently working on or learning. We want to see that you are active in this space right now, not that you took a course two years ago.
Applications with completed videos get reviewed first. Bonus points if you also include a GitHub, portfolio, or demo link.
The Role
You will be the technical owner of our AI agent ecosystem. We currently run 30+ automated skills and agents through Claude (Anthropic), OpenClaw, and various API integrations (HubSpot, ---------- , Google Workspace, RingCentral). Your job is to deploy, maintain, improve, and scale these agents — and build new ones as the business needs them.
This is NOT a traditional software engineering role. You will be working with AI agent frameworks, prompt engineering, API integrations, automated email systems, CRM pipeline automation, and server deployment. You are essentially building AI employees that do the work of real staff members — generating reports, moving deals through pipelines, sending follow-up emails, updating dashboards, and surfacing insights to managers automatically.
You need to be comfortable figuring things out fast in a rapidly evolving AI tooling landscape.
what we need done.
Pipeline & Lead Hygiene (Sales / BCO)
• Auto-flag inactive leads with no activity after 7 days and alert the manager
• Auto-flag leads with fewer than 3 contact attempts within 14 days
• Enforce the 400-lead cap and remove overflow automatically
• Check if a lead is already a client or on the DNC list before it ever gets dialed
• Distribute new leads evenly across setters via round-robin logic
• Run the weekly BCO city report and flag cities below Alliance minimums
• Assign leads as tasks to BCO Scouts in bulk every Monday
HubSpot Logging & Deal Creation (Sales)
• Create callback tasks with exact date, time, and context notes
• Set appointment booked disposition and trigger the thank-you email automatically
• Run the cleaning calculator and populate the deal price field
• Associate contact, call, task, and meeting to the deal in a single action
• Create the pre-presentation task and assign it to the correct presenter
• Enroll prospects in the email sequence when they request email-only follow-up
• Flag any HubSpot record not updated before end of shift
Daily Performance Monitoring (Sales / BCO / CS)
• Alert the manager if a setter hasn't started dialing by 9:05 AM
• Track daily dials and send a pace alert if they are behind by midday
• Calculate and report daily talk time per rep
• Generate an end-of-day KPI summary per setter automatically
• Measure rep downtime — flag any gap over 5 minutes between calls
Ticket Management (Client Services)
• Pull the client checklist and verify whether the complaint is in scope before any action is taken
• Format the ticket title correctly on every ticket, every time
• Determine the correct ticket type (New Concern, R1, Move to BCO) by checking 3-month client history
• Create and pin the Master Note with all required fields auto-populated where data exists
• Auto-tag Edgar, Peter, Taylor, and the assigned handler on every note
• Send the templated client acknowledgment email immediately on ticket creation
• Schedule the follow-up email at 8 AM the morning after the correction clean
• Send the second follow-up if the client does not respond (Template F)
• Auto-open a service ticket when a survey score comes back at 5 or below
• Send the R1 Warning Template the ---------- nt a second-strike ticket is created
• Block ticket closure if the closure checklist is not complete
• Flag any ticket open longer than 7 days
• Track monthly attrition and alert when it approaches $6K
BCO Pipeline Operations
• Move ticket stages automatically as milestones are hit
• Calculate profit splits based on building type and billing amount
• Check Alliance issue history and active contract count before any match is made
• Pin verification notes and tag the right people on every one
• Send the Welcome Email on approval and the contract via Google Signature
• Download signed contracts and file them to the correct Drive folder automatically
• Check inspection report submission by 8 AM after every first clean and alert if missing
• Auto-create a BCO ticket the ---------- nt an Alliance partner gives up a building
• Send email and SMS outreach sequences automatically on no-answer
HR & Hiring
• Write the three screening questions for each new job ad
• Review ---------- for new applicants as they auto-populate
• Send reference request outreach to candidates via email
• Create Google Meet links with correct titles and send confirmation emails
• Run the Claude Analyzer after every interview, save the PDF, create the Drive link, and paste it into ----------
• Log manager interview notes from email into the candidate's ---------- record
• Run the end-of-day board audit and flag any fields that were not updated
Performance Documentation (All Departments)
• Pull the full performance history for any employee before a coaching conversation
• Check whether a written expectation was ever actually sent
• Check whether a deadline was ever actually set
• Check Trainual for training completion before any performance action is taken
• Draft the post-coaching confirmation email from structured notes
• Flag when an escalation threshold is hit (3rd issue, missed deadline pattern)
• Compile the full escalation documentation package before a termination
Core Infrastructure (Ongoing)
• Deploy and manage OpenClaw AI agents on Linux VPS and Docker — ensure 24/7 uptime
• Build, test, and improve agent skills across sales, service, BCO/HR, and executive reporting
• Maintain and extend integrations with HubSpot, ---------- , Google Workspace, and RingCentral
• Build and maintain HTML dashboards and reporting tools used by managers daily
• Manage the agents' knowledge layer — SOPs, KPIs, org structure, and process documentation
• Train non-technical tea ---------- mbers on browser-based tools and new systems
• Troubleshoot agent failures, API errors, and integration breakdowns — fix them fast
• Document everything you build so it is not locked in your head
Required Skills & Experience
Must-Have:
• 2+ years working with Node.js / JavaScript in a production environment
• Experience deploying and managing applications on Linux servers (Ubuntu preferred)
• Comfortable with Docker, SSH, and basic DevOps (deploying and maintaining agents on VPS instances)
• API integration experience — REST APIs, webhooks, and OAuth flows
• Experience with Google Workspace (Drive, Gmail, Calendar, Sheets) including API-level integration
• Experience building automated email workflows or CRM automation (HubSpot, Salesforce, or similar)
• Strong English communication — you will work directly with non-technical managers who need clear explanations
• Ability to train non-technical users on browser-based tools — patience, clear documentation, screen recordings
• Self-directed and proactive — you do not wait to be told what is broken, you find it and fix it
Strong Preference:
• Experience with AI agent frameworks (OpenClaw, LangChain, CrewAI, AutoGen, or similar)
• Prompt engineering skills — writing effective system prompts, managing context windows, and building knowledge bases for LLMs
• Experience with HubSpot API (deals, contacts, tickets, pipelines, workflows) and ---------- API
• Experience building agents or bots that automate multi-step business processes (not just chatbots)
• Familiarity with vector databases (ChromaDB, Pinecone) and RAG architectures
• HTML/CSS for building dashboards and reports
• Experience working with a small team remotely across time zones
Nice-to-Have:
• Python scripting (some of our automation uses Python)
• Experience with Pipedream or serverless deployment
• Familiarity with Anthropic's Claude API or OpenAI API
• Background in business process automation (not just coding — understanding why we automate)
• Experience with ---------- , Zapier, or n8n workflow platforms
How to Apply
Compensation & Growth
• $1,200 – $2,400 CAD/month depending on experience and skill match
• Performance reviews at 30, 60, and 90 days with potential raise based on results
• Long-term role — we are building a permanent AI operations layer, not a one-off project
• Autonomy — you own the technical stack. If something should be done differently, you tell us
• Growth opportunity — as One Janitorial scales, so does the AI infrastructure and your role with it