Senior Make.com Integration Architect – PMS, Smart Lock & AI Workflow Automation

Hire a Pro Hire Help
1

We are building a production-grade automation system for a short-term rental portfolio (25 units, scaling to 100+).

We are looking for a senior Make.com integration architect to design and implement a modular automation system integrating:

• PMS (Nowistay or similar)

• TTLock API (smart lock events + PIN generation)

• WhatsApp Cloud API

• Google Sheets (operational database)

• Google Forms (task execution interface with pre-filled parameters)

• OpenAI Vision API for automated image analysis

Scope of work includes:

  1. Event-driven task creation upon checkout

  2. Dynamic Google Form link generation using pre-filled parameters

  3. Processing form submissions (checklists + image uploads)

  4. Sending images to OpenAI Vision and routing based on JSON response

  5. Updating PMS room status based on QA results

  6. WhatsApp notifications for staff and operations

  7. Clean, scalable Make scenario architecture (modular design)

Requirements:

• Advanced Make.com scenario design experience

• Strong HTTP/API integration knowledge

• Experience handling webhooks and structured JSON

• Clean logging within Google Sheets

• Ability to build retry/fallback logic for failed API calls

• Clear documentation and handover

This is a system design and architecture role, not a simple automation setup.

Please share:

• Examples of complex Make integrations

• Description of your API-heavy automation work

• Your hourly rate

2 Likes
2

Hello @Fahad2 , welcome to make.com community, I have worked and have experience with Make.com and l will love to collaborate with you on this you can schedule a call Here and you can checkout my upwork profile Here, for my pastworks and certifications

3

Hi Fahad, Welcome to the community!!

Your project scope is well structured and clearly positioned at architecture level, not just automation setup, which is great to see.

I work extensively with API heavy Make.com systems involving webhook orchestration, modular scenario design, WhatsApp Cloud API, Google Sheets as an operational layer, and OpenAI based routing with structured logging and fallback logic. Production stability and scalability are always my priority.

A few things I would love to better understand:

• How do you envision the system behaving when an API fails or returns incomplete data?

• What does a fully successful automation flow look like to you from checkout to QA completion?

• Are you planning to scale this beyond 100 units in phases or aggressively?

Feel free to reply directly to my email so we can discuss more clearly: [email protected]

[email protected]

You can also view my portfolio here:

Book a call if that works better: https://cal.com/folafoluwa-olaneye-osrofp/30min

Or continue securely via my Fiverr workspace where I manage freelance projects:

http://www.fiverr.com/s/qD1jlXZ

Looking forward to hearing from you.

Best regards,

Folafoluwa Stephen

Flowedge Digitals

4

Hey, this is exactly the kind of production-grade Make.com architecture I build.

I can design a modular system that listens to checkout events from your PMS, generates pre-filled Google Form links for task execution, processes submissions (checklists + photos), runs OpenAI Vision analysis, updates room status back in the PMS, and sends WhatsApp notifications to staff/ops all with clean logging in Google Sheets and solid retry/fallback handling.

I’ve built API-heavy Make scenarios with webhooks, structured JSON routing, and multi-step orchestration where downstream systems (like AI enrichment or external APIs) can be slow or fail so I’m comfortable designing for reliability, idempotency, and scale from 25 units to 100+.

:page_facing_up: Check my profile: https://www.upwork.com/freelancers/farhana401

:telephone_receiver: Book a quick call:Calendly - Automaxion

5

Fahad:

We’re a Make Gold Partner specializing in production-grade automation architecture for complex operational systems.

This is absolutely a system architecture engagement, not a simple scenario build. Event-driven triggers from a PMS, smart lock telemetry, structured JSON routing, image analysis via OpenAI Vision, dynamic form generation, QA logic, status updates, and WhatsApp notifications — all wrapped in modular, retry-safe design. That’s exactly what we do.

Quick API confirmation:

  • TTLock API — Open API available. We’ll connect via secure HTTP for lock events and PIN generation.

  • WhatsApp Cloud API — Meta’s Graph API, fully supported through HTTP modules.

  • Google Sheets — Native Make integration.

  • Google Forms — We can generate pre-filled links dynamically and process submissions via webhooks.

  • OpenAI Vision API — HTTP calls to OpenAI endpoints, fully supported.

  • PMS (Nowistay or similar) — We’ll confirm exact API specs during discovery, but based on initial review it provides API access.

We just finished a similar build for another client — Bluetooth lock events, real-time access logs, operational database updates, downstream workflow triggers, and structured audit logging. The architectural patterns you’re describing are very much in our wheelhouse.

For a 25-unit portfolio scaling to 100+, we’d design this as:

Modular scenario architecture with isolated components for:

  • Lock event processing

  • Checkout triggers

  • QA workflow + image analysis routing

  • PMS status updates

  • Staff notification logic

This keeps it scalable, debuggable, and production-safe as you grow.

We handle authentication, token refresh, JSON parsing, retry/fallback logic, and clean logging as part of the build — not as optional extras.

We also fully document everything we do (sample attached).

Next steps:

Book a discovery call here: https://4spotconsulting.com/30

I’ll DM you separately with our rates.

Keep Automating,
Jeff Arnold
Founder & President, 4Spot Consulting

Screenshot 2026-02-23 at 9.48.30 AM
Screenshot 2026-02-23 at 9.48.30 AM1590×584 128 KB

Screenshot 2026-02-23 at 9.48.40 AM
Screenshot 2026-02-23 at 9.48.40 AM1590×584 100 KB

Screenshot 2026-02-23 at 9.48.52 AM
Screenshot 2026-02-23 at 9.48.52 AM1590×695 134 KB

Screenshot 2026-02-23 at 9.49.04 AM
Screenshot 2026-02-23 at 9.49.04 AM1590×832 152 KB

4000189 - [*ACTIVE]- LinkedIn Outreach - Part 1 - When a New ID added in Airtable - Visit profile - Update status_Redacted.pdf (1.1 MB)

6

Hi Fahad — this is exactly the kind of Make.com work I handle at Hashlogics (modular, API-heavy, production architecture rather than a one-off scenario)

I’ve built automation systems where Make sits as the orchestration layer between operational platforms, messaging channels, and AI services, with strict routing, retries, and clean handover. Your stack and flow make sense, and the key here is to design it as modular scenarios (not one giant scenario) so it scales from 25 units to 100+ without becoming hard to maintain

My approach would be to split this into core modules: checkout/event intake, task creation + Google Form prefill generation, form submission processing (checklist + image uploads), Vision analysis + JSON-based QA routing, PMS status update, and WhatsApp notifications. I’d also keep a central run log in Google Sheets (job ID / unit / task type / status / retries / final outcome) with idempotency keys so duplicate webhook events don’t trigger duplicate tasks or PIN actions

For reliability, I’d add retry/backoff for PMS/TTLock/WhatsApp API calls, exception queues for failed image analysis or missing uploads, and a reconciliation job (e.g., scheduled health checks for tasks stuck in-progress). That is usually what separates a working demo from a system your ops team can trust daily

I’ve done similar API-heavy automation work involving webhook-driven workflows, structured JSON routing, AI-based decisioning, and operational notifications with proper logging/monitoring. I can also provide a clean documentation + handover pack (scenario map, data model, retry behavior, and support SOP)

My hourly rate is $50/hr

If you want, send over the PMS API docs (or confirm Nowistay version), TTLock endpoints you plan to use first, and a sample QA checklist/form structure, and I can outline the exact scenario architecture before we start

:star: Reviews: Clutch
:globe_with_meridians: Portfolio: Case Study

:date: Book a call: Calendly

7

This is a well-scoped project with some interesting technical challenges, particularly around the TTLock integration and the OpenAI Vision QA pipeline.

**What caught my attention:**

The event-driven checkout flow into dynamic Google Form generation is smart — it means your cleaning staff gets a pre-filled checklist without needing a separate app. The real complexity is in the vision QA loop: image upload, OpenAI analysis, pass/fail routing, and PMS status update all need to happen reliably with proper error handling.

**Relevant experience:**

I build API-heavy automation systems daily. A recent project involved integrating Amazon Ads API (OAuth2), Keepa API, and ClickUp API into a multi-agent intelligence system with structured JSON processing, conditional routing based on AI analysis results, and automated Slack notifications — very similar pattern to your checkout-to-QA pipeline.

**How I’d approach your key technical challenges:**

- **TTLock integration:** HTTP module with OAuth token management, webhook listener for lock events, and a PIN generation scenario that fires on confirmed booking. The critical part is handling the token refresh cycle reliably since TTLock tokens expire.

- **OpenAI Vision QA:** Image URLs from form submissions sent to Vision API with a structured JSON schema enforcing pass/fail + specific issue categories. Router splits based on the JSON response — pass updates PMS status directly, fail triggers WhatsApp alert to operations with the flagged images attached.

- **Scalability for 100+ units:** Modular scenario design from day one — each workflow (checkout trigger, form processing, QA analysis, notifications) as a separate scenario connected via webhooks. This means you can update one without touching others, and operations limits scale linearly.

- **Error handling:** Every external API call gets a retry/fallback path. Failed TTLock calls queue to a Google Sheet “retry log” with full context. No silent failures.

**Questions before scoping:**

1. Is Nowistay confirmed as the PMS, or are you still evaluating? API availability varies significantly between PMS platforms.

2. For TTLock — are the locks already deployed, and do you have API credentials set up?

3. How many images per QA inspection on average? This affects OpenAI API costs and scenario operation counts.

4. Do you need multi-language WhatsApp notifications for cleaning staff?

**About me:**

- Founder of Evara AI (incubated at IIT Bhubaneswar) — custom automation and AI systems

- 1.5+ years of production API integrations (REST, OAuth2, webhooks, structured JSON)

- Time zone: IST (UTC+5:30), available 20+ hours/week

- Happy to share hourly rate after understanding full scope

I’d be glad to walk through the architecture on a quick call. Reach me at: [email protected]

8

Hi,

This is the kind of system that should be treated as operations infrastructure, not as a loose collection of automations. For PMS events, smart lock data, WhatsApp messaging, forms, and vision analysis to work reliably together, the key is event design, idempotency, traceability, and clear fallback paths when an external API is delayed or incomplete.

The right entry point is usually one narrow production path, for example checkout to task generation to form execution to image review, and then expanding once the failure handling is solid. If you are still hiring, I would be comfortable starting there as a paid implementation.

9

Hi there,

I’ve reviewed your requirements for the Short-Term Rental (STR) portfolio, and I’m ready to design a scalable, production-grade architecture for your 25+ units. I specialize in building event-driven systems where reliability, logging, and error handling are the core, not an afterthought.

How I will approach your system design:

  • Modular Architecture: I’ll split the logic into “Micro-scenarios” (Trigger, Logic, Action) connected via Webhooks/Data Stores. This ensures that a failure in the OpenAI Vision module doesn’t stop the TTLock PIN generation.

  • Deep API Integration: I’m experienced in handling HTTP/OAuth2 for property management systems and hardware APIs (like TTLock). I’ll implement Retry/Fallback logic to ensure that if a PMS or Lock API is momentarily down, the task is queued and retried, not lost.

  • AI-Powered QA: I’ve implemented OpenAI Vision flows that parse images into structured JSON. I’ll ensure the “Cleaning Checklist” images are accurately analyzed to trigger “Room Ready” status in your PMS automatically.

  • Operational Transparency: I’ll build a centralized Audit Log in Google Sheets, tracking every webhook and API response for easy debugging and staff performance monitoring.

Experience & Work:

  • Complex Integrations: I’ve built multi-stage lead factories and ERP-style systems (30+ screens/states) involving GHL, PostgreSQL, and custom API middleware.

  • API-Heavy Work: Expert in JSON transformation, complex filtering, and state management within Make.com and n8n.

  • Hourly Rate: $50/hr (depending on the final scope and technical complexity).

Portfolio: https://mikedevai.netlify.app/ Direct Contact: @hely_chatbots (Telegram)

10

I would start this as one production checkout-to-room-status slice, then expand once the state model is proven.

For the first Make build, I would map the PMS checkout event into a task row, generate the pre-filled Google Form for the cleaner or inspector, process the form/image uploads, run the OpenAI Vision check only when required, update the PMS room status, and send the WhatsApp/ops notification from the same audit trail. TTLock PIN changes should be idempotent and logged separately so a rerun cannot create conflicting access events.

TinyOps Studio can quote a fixed pilot for that first slice with the data contract, run log, retry rules, and handoff notes included. If it works cleanly for one unit/task type, the same pattern can be cloned across the rest of the portfolio.

If this is still open, DM me the first PMS event sample and the room-status states you want, and I can scope the pilot.