Description
Summary We run a small operations consultancy building light automation for local service businesses (clinics, salons, studios). We need a freelancer to design and build one working sample pipeline for appointment reminders and no-show recovery, using a compliant messaging channel (WhatsApp Business API or SMS gateway) integrated with a calendar/booking system (Google Calendar, Calendly, Acuity - your recommendation is welcome, as long as it is a real, working integration). This is a paid trial engagement structured across multiple milestones. Total engagement value across the full scope is $150-300 over multiple milestones. We are starting with a small, clearly-defined first milestone to confirm fit and execution quality. Milestone 2 onwards is conditional on Milestone 1 acceptance - there is no guaranteed total beyond Milestone 1; later milestones are scoped and funded only if the first is accepted. SCOPE OF WORK (Milestone 1 - this posting) Build a working sample automation for one representative business (a demo/sandbox clinic, salon, or studio - your choice) that includes: 1. Calendar integration: pull upcoming appointments from a calendar/booking system. 2. Reminder automation: send automated reminders via a compliant channel - WhatsApp Business API (official/sandbox) or an SMS gateway following carrier/platform rules (e.g., Twilio). No unofficial/grey-market WhatsApp tools. 3. No-show recovery flow: a follow-up sequence triggered when an appointment is marked no-show (rebooking prompt / short recovery sequence). 4. Test data handling: we will provide a test calendar dataset. It intentionally contains some data irregularities (a normal part of our QA process). Your deliverable must explicitly flag any irregularities you encounter (timezone mismatches, duplicate/overlapping bookings, malformed entries) rather than silently normalizing, dropping, or working around them. We value an honest, clearly-flagged issue report over a pipeline that looks clean but quietly papers over data problems. DELIVERABLES (target 5-7 days from start) - A working, runnable pipeline (code + configuration): calendar pull to reminder send to no-show follow-up. - Configuration documentation: setup steps, placeholder credentials/env vars (no real secrets), and how to adapt it to a different business. - A step-by-step AI-vs-manual time log per build step (granular and honest - a required deliverable). - A written flag report listing every data irregularity found in the provided test calendar, what you did, and why. ACCEPTANCE CRITERIA (all required) - Pipeline runs end-to-end against the provided test calendar and produces reminder + no-show messages to a sandbox/test contact. - Messaging channel is compliant (official WhatsApp Business API or standard SMS gateway) - no grey-market tools. - Documentation is reproducible by a third party. - Time log present and itemized by step. - All test-data irregularities explicitly flagged in writing (none silently fixed or omitted). BUDGET Milestone 1 (this posting): $25-30 fixed price, paid on acceptance. Total potential engagement value if we continue: $150-300 across multiple milestones - milestone 2 onwards is scoped and funded only after Milestone 1 is reviewed and accepted. No guaranteed total beyond Milestone 1. TO APPLY, please answer briefly in your proposal: 1. Describe a time you integrated a calendar/booking system with a messaging channel (WhatsApp Business API, Twilio SMS, or similar). What compliance considerations (opt-in, carrier rules, template approval) did you handle? 2. If you were processing appointment records and noticed conflicting timezones or duplicate time slots, what would you do and why? 3. Walk us through how you would structure a no-show recovery message sequence (timing, number of follow-ups, tone) for a small clinic or salon. 4. Have you used the official WhatsApp Business API (Cloud API or on-premise) before, or only third-party unofficial tools? Please be specific. 5. This milestone requires a granular, honest AI-assisted vs. manual time log per step, even where AI did most of the work. Are you comfortable providing that? Please keep all work and communication on Upwork.