Multiagent OmniFlow for Routing
Multiple Search Indexes in Data Cloud
Carvilla Multiagent Architecture
Slack messages as Direct Message to Test Driver
Slack messages to the channel
CarVilla Website (single pager) in Experience Cloud

Project Story – Autoforce: The Future of Automotive Experiences

Inspiration

We were inspired by the fragmented customer experience in the automotive industry. From browsing vehicles to booking test drives and handling service requests, most car buyers still navigate disconnected channels, long wait times, and inconsistent support.

We asked ourselves:
What if AI-powered agents could orchestrate a smarter, seamless customer journey, end-to-end, and do it all on Salesforce?

That led to Autoforce: a network of intelligent agents built with Agentforce, designed to modernize the car buying and ownership experience.

What it does

Autoforce is a multi-agent orchestration solution built entirely on Salesforce.

Master Agent answers customer questions using PDF and Knowledge Base ingestion.
Test Drive Agent schedules, reschedules, and cancels test drives with real-time Slack and email notifications.
Service Agent logs vehicle issues, updates service requests, and escalates to live reps using Omni-Channel.

Each agent is specialized but connected, passing context and collaborating for a seamless user journey.

Live prototype:
https://carvilla-dev-ed.develop.my.site.com/carvilla

How we built it

We used an all-native Salesforce stack:

Agentforce for intelligent agent orchestration
Flows and Omni-Channel for contextual routing and escalations
Slack API for real-time internal rep notifications
LLM & PDF ingestion for handling general product inquiries
Experience Cloud to host the front-end user experience

Each interaction is designed to be modular, agent-led, and extensible, mirroring real dealership operations.

Challenges we ran into

Designing seamless agent-to-agent context handoff
Ensuring real-time Slack alerts didn't break flow or consistency
Handling natural language inputs in a structured, context-aware way
Integrating PDF-based product data into accurate agent responses

Accomplishments that we're proud of

We built a fully functional, end-to-end automotive agent experience, all within Salesforce.
Achieved natural, human-like query handling from knowledge + PDFs.
Delivered a working test drive scheduling with Slack alerts and customer email confirmations.
Maintained architecture clarity while building a multi-agent framework.

What we learned

Agentforce is incredibly powerful when paired with Flow and Omni-Channel
The future of automation is not one AI assistant, it’s many, working together
Slack, when used well, is a perfect complement to AI-driven support
Designing for business outcomes, not just features, leads to more impactful experiences

What's next for Autoforce (The Future of Automotive Experiences)

Adding support for connected vehicle data (e.g., diagnostics, recalls, remote services)
Personalizing agent behavior using dynamic segments or Data Cloud insights
Extending orchestration to post-sale journeys like warranty, insurance renewals and trade-ins
Enabling voice-powered agent interaction in dealership kiosks or in-car systems

Built with passion by
Team Autoforce – Anil, Vishal, Ranjeet, Narendra & Honey