Where Love is headed
Connect your CRM to your AI agent.
Imagine asking the agent you already use — Claude, Codex, ChatGPT, or one you built yourself — to run your book of business, and having it actually do the work.
This is the direction we are building, shaped with design partners as it lands. It is not a shipped feature.
Why it matters.
Connecting a CRM to an AI agent means an assistant can securely read your customer, job, and billing records and take action in them — instead of a person clicking through screens to get the same result.
Legacy field service software was built for humans operating a screen. Every record lives behind a form, every workflow behind a click. An agent can describe that work, but it cannot do it.
Love is being built the other way around: a system of record that is also a system of action, addressable by the agents your team already trusts. That is the direction — not a feature we are claiming today.
The roadmap, in three steps
How it would work.
Connect your agent
Authorize the agent you already use. Love issues it a scoped connection to your workspace — no exports, no scraping, no brittle integrations to maintain.
It reads your business
Customers, properties, service history, routes, estimates, invoices, and conversations become structured context the agent can query directly.
It does the work
Booking, rescheduling, follow-ups, collections, reporting — the agent performs the work inside Love, and every action lands on the record.
In practice
What you would ask it.
- Which accounts are at risk of churning this month?
- Reschedule tomorrow's route around the storm and text every affected customer.
- Draft follow-ups for every unsold estimate over $2,000.
- Who on the crew is running behind today, and what should we move?
- Summarize why we lost the last twenty cancellations.
- Collect on every invoice past 30 days and tell me what's left.
Design principles
Powerful, and still accountable.
Scoped permissions
Every agent connection gets only the access it needs — read-only, a single module, or a specific set of actions.
A full audit trail
Every read and every write attributed to the agent that made it, reviewable by the owner at any time.
Your data stays yours
Your book of business belongs to you. Connecting an agent should never mean handing your customer list to someone else.
These are the principles guiding the build, stated as intent rather than as certifications or guarantees available today.
Frequently asked
Questions, answered.
- What does it mean to connect a CRM to an AI agent?
- It means an AI assistant can securely read your CRM data — customers, jobs, routes, invoices, history — and take action in it, instead of a person clicking through screens. The agent asks the CRM questions and performs work through it.
- Which AI agents would Love connect to?
- The direction is agent-agnostic: assistants like Claude, Codex, and ChatGPT, plus custom or in-house agents a company builds itself. Love is designed to be the system of record and system of action those agents work through.
- Is this available today?
- No. This is roadmap direction being shaped with early design partners. Nothing on this page is a shipped feature or a dated commitment.
- How would data stay safe?
- The design principles are scoped permissions per agent, explicit approval for sensitive actions, a complete audit record of everything an agent does, and customer data that remains owned by the business.
- How do I get access when it rolls out?
- Request early access on this page. Design partners are the first to get agent connections as they land.
Help us build the bridge.
Design partners shape what agents can do inside Love — and get connections first as they roll out.
Read why we exist