← Live-Feed

AI Automation & Fullstack Developers — Build DominOS for Domino’s Switzerland

Budget: $25.0 - $40.0 HOURLY / NOT_SURE ⭐ 0.00 (0) Switzerland

Bevorzugte Qualifikationen

  • Erfahrung: Experte
We are looking for two to three developers for a long-term collaboration building DominOS, the AI Automation Engine for the Domino’s master franchise in Switzerland. Complementary individual applicants are welcome: agentic/backend engineering and technical leadership, n8n/integrations, and fullstack/product interfaces. Nobody needs to cover everything alone. Our goal is to automate operational and administrative work, reduce manual effort, errors and costs, and gradually reduce the headcount required to run the business. Staffing and other sensitive decisions remain with responsible people. COMPENSATION USD 25–40 per hour, per developer. Long-term collaboration starts with a paid pilot; scope and acceptance criteria are agreed before work begins. HOW WE BUILD Idea → rapid prototype → real-world test → measurable validation → controlled production rollout → continuous improvement. We want people who discover opportunities, not just complete tickets. Research new approaches, propose focused experiments, test their practical value and integrate what works into a coherent system. Store automation, RFID inventory, shift planning, content workflows, HR onboarding/offboarding and financial document retrieval are examples, not a closed roadmap. TECHNICAL DIRECTION • Hermes/agentic-first: reusable, versioned Skills, tool calling, structured outputs and evaluations. • n8n: defined integrations, events, approvals, waiting states and error handling. • Backend/frontend: shared data models, APIs, roles and usable interfaces. • Existing connectors first; build new ones only where needed. Models are interchangeable tools. Permissions, calculations and explicit rules remain deterministic. 24/7 means monitored system operation, not round-the-clock developer availability. FIRST PAID ASSIGNMENT: RFID MVP We provide handheld RFID readers and RFID printers for development. Agree scope and acceptance criteria before starting; this is not an assignment to implement all of DOMPILOT. 1. Scan delivery barcodes and reconcile against delivery notes; flag missing data rather than guessing. 2. Print and encode unique RFID labels linked to product, quantity and available batch/best-before data. Printing alone must not confirm goods receipt. 3. Document discrepancies and prepare evidence-backed supplier claims; sending follows an agreed approval policy. 4. Label goods during food preparation and track carton-to-prepared-unit transitions without double-counting stock. 5. Support daily handheld inventories: deduplicate reads, reconcile actual versus expected stock and flag discrepancies. 6. Deliver a clear manager interface for receiving, stock, counts, discrepancies and claim status. Handle discarded labels, print failures, connection interruptions and chilled/frozen label conditions. Inventory changes through defined state transitions, not arbitrary reads. Validate against manual counts, time spent, errors, usability and recovery. Forecasting and NFC check-in may follow; Pulse write access is outside this initial scope. SKILLS AND DELIVERY Practical Python and/or TypeScript, APIs, relational databases, n8n, webhooks, OAuth/MCP, hardware SDKs/device APIs for the RFID work, Git, tests, reviews, Docker/Linux, deployment, monitoring, backups and rollback. Build idempotency, retries, timeouts and reconciliation into workflows. Enforce least privilege and prompt-injection protections outside the model. Hermes experience is a plus; willingness to learn is required. We use a shared backlog, our repositories and documentation. Julian and Nico jointly own strategy, priorities and budget. Nico is your direct contact for business acceptance; developers own technical quality. A technical lead owns architecture and reviews. Expect short delivery cycles, early demos, written updates and clear handovers. Working windows/support are explicitly agreed; reliability matters more than location. Done means reproducible setup, normal/failure tests, verified permissions, confirmed target-system outcomes and documentation another developer can use. Production additionally requires monitoring, rollback and a named owner. PLEASE APPLY WITH • Your strongest role and 1–2 real projects, specifying your contribution. • A production failure and your sustainable fix. • A short assessment of RFID integration risks; no unpaid detailed planning. • One automation idea for a multi-location business and a quick value test. • Time zone, location, start date, reliable weekly availability and proposed rate. Clear English communication/documentation required; German is a plus. We start with a paid pilot. Confidentiality, usage rights and access are agreed before work. Code, Skills and workflows stay in our accounts. Nico & Julian
Auf Upwork öffnen

AI proposal draft

Generate a short cover letter for this job. Edit before sending.

Sign in to generate an AI proposal draft.

Anmelden