2 ready-to-build workflows

AI agent workflows for Automotive & dealerships

Book test drives, chase service due dates and answer parts enquiries — before a rival dealer calls back.

Each recipe below is expressed only in VegaDūta's real workflow building blocks — Trigger, Agent, Knowledge, Condition, Approval, Output, Tool (MCP), Voice, Device, Loop, Code and Parallel — so it maps 1:1 to something you can assemble in the Workflow designer. Consequential actions always pass a human Approval step.

1. Test-drive booking & follow-up, end to end

Enquiries land by WhatsApp and missed calls at all hours. Someone has to check which variant is in stock, offer a slot, then remember to follow up — so leads go cold and showroom slots sit empty.

Trigger: WhatsApp / Call (enquiry)

How the workflow runs

  1. Trigger

    Buyer enquires. A WhatsApp or Call trigger starts the flow on any inbound enquiry about a model or a test drive.

  2. Agent

    Qualify the interest. An Agent node captures the model/variant, budget and preferred timing, and confirms the buyer's city/dealership.

  3. Knowledge

    Check live inventory. A Knowledge node grounds the reply in your current stock and demo-car availability — so it never offers a variant you don't have.

  4. Output

    Offer + book a slot. An Output node proposes test-drive slots and confirms the chosen one on WhatsApp.

  5. Tool (MCP)

    Write to the DMS. A Tool (MCP) node records the booking and the lead in your dealer management system — no first-party DMS connector, MCP calls yours.

  6. Output

    Reminder + follow-up. An Output node sends a day-of reminder, then a post-drive follow-up to move the buyer toward a quote.

Channels & connectors

  • WhatsApp
  • Voice / Call
  • Knowledge base
  • Tool (MCP → DMS)

Outcome

Every enquiry is qualified, offered a real in-stock slot, logged, and followed up — without a salesperson tied to the phone.

Why it helps

Speed-to-lead and a booked slot are what convert walk-in intent; doing both instantly beats a callback hours later when the buyer has moved on.

Build spec

1 agent — 1 sales agent handles the conversation; inventory grounding, the DMS write and the reminders are workflow nodes.

System prompt (paste-ready)

You are a dealership sales assistant for {{dealership}}. From the buyer's message, capture the model/variant of interest, their budget band and preferred timing, and confirm their city/dealership. Only offer a test-drive slot for a variant that Knowledge confirms is in stock or has a demo car available — never promise a variant you can't confirm. Work in the buyer's language. Reply warmly and briefly. Once the buyer picks a slot, hand off to the booking step; do not claim the booking is confirmed until it is written to the DMS.

MCP connectors

  • Dealer management system (DMS) — via a Tool (MCP) node (no first-party connector)
  • WhatsApp — send channel

Built-in tools

  • knowledge_search (live inventory + demo-car availability)

Guardrails

  • Grounded-only: slots are offered only for variants Knowledge confirms in stock — never an unavailable model
  • The booking is not treated as confirmed until the Tool node has written it to the DMS
  • Scope-locked to the enquiring buyer and their chosen dealership — the agent can't read or book other stores

Output & delivery

The agent proposes real in-stock slots → the buyer confirms one on WhatsApp → a Tool node records the booking + lead in the DMS via MCP → Output nodes send a day-of reminder and a post-drive follow-up toward a quote.

2. Proactive service reminders that book themselves

Service due dates are tracked in a spreadsheet and reminders are called manually — so cars miss their service, the workshop bay sits idle, and post-service feedback is never collected.

Trigger: Webhook (scheduled, service-due run)

How the workflow runs

  1. Trigger

    Service-due run. A scheduled Webhook trigger evaluates vehicles approaching their service date each morning (fed from your DMS via Webhook).

  2. Loop

    Per due vehicle. A Loop node iterates the due list so each owner gets a tailored reminder.

  3. Agent

    Draft the reminder + slot options. An Agent node writes a friendly reminder with the car's due item and offers workshop slots.

  4. Output

    Reach the owner. An Output node sends it on WhatsApp/SMS and lets the owner pick a slot in one tap.

  5. Tool (MCP)

    Book the workshop bay. A Tool (MCP) node writes the confirmed booking back to the DMS / workshop scheduler.

  6. Output

    Post-service feedback. An Output node sends a short feedback ask after the service is marked complete.

Channels & connectors

  • Webhook
  • Loop
  • WhatsApp
  • SMS
  • Tool (MCP → DMS/scheduler)

Outcome

Due vehicles are reminded and booked automatically, bays fill up, and feedback arrives without service advisors chasing it.

Why it helps

Consistent, timely reminders are what bring cars back in — the manual process skips exactly the customers who forget, which is most of them.

Build spec

1 agent — 1 agent drafts each reminder inside the Loop; the scheduled trigger, the loop, the booking write and the feedback ask are workflow nodes.

System prompt (paste-ready)

You are a service-reminder assistant for {{dealership}}'s workshop. For each vehicle you receive (owner name, model, registration, due service item, due date), write a short, friendly reminder in the owner's language that names the specific due item and invites them to pick a workshop slot. Do not invent service items, prices, or offers beyond what the record provides. Keep it to a couple of lines and one clear call to action.

MCP connectors

  • DMS / workshop scheduler — via a Tool (MCP) node
  • WhatsApp + SMS — send channels

Built-in tools

  • http_request (pull the vehicle's due item / service history)

Guardrails

  • Idempotent per vehicle per service cycle — a re-run of the daily job never double-reminds the same owner
  • Scope-locked to the vehicles in the due list for this run — no messaging owners who aren't due
  • A slot is only treated as booked once the Tool node writes it back to the scheduler

Output & delivery

The Loop iterates the due list → the agent drafts a tailored reminder → an Output node sends it on WhatsApp/SMS with one-tap slot pick → a Tool node books the bay in the DMS/scheduler via MCP → after the service is marked complete, an Output node sends a short feedback ask.

Related industries

Build one of these in minutes

The sandbox gives you a live agent workspace — no account, no card. Or head back to the full catalogue and compare patterns across every industry.

Try this in the sandbox — no signupBrowse all industries
See how VegaDūta compares to n8n, Dify and BotpressWhy VegaDūta's architectureEstimate your WhatsApp API costs