LinkedIn InMail is an example of an agent your team can build in Sylvie: it reads a prospect's profile and recent activity, then writes the subject line and short body of a one-to-one InMail that gives that specific person a reason to reply. It runs on your brand brain, so the angle comes from your real positioning, the case study that matches their industry, and the voice of the person whose name is on the message. Sylvie does not ship it pre-built: your team assembles it on your own memory, so every credit you spend goes out sounding like you.
Sylvie does not ship this agent pre-built. It is an example of an agent your team can build with Sylvie, powered by your brand brain, so it runs your workflow the way you actually run it.
LinkedIn InMail is an agent you assemble inside Sylvie, not a feature you switch on. Give it a prospect, a list or a saved Sales Navigator segment and it drafts each message one at a time: a subject line short enough to survive a phone screen, three or four lines connecting what that person is dealing with to something you have actually done, and a single low-friction ask.
InMail is expensive attention. Credits are finite, the recipient never asked to hear from you, and your message lands next to messages from their own team. That is why the generic 'quick question' template dies on arrival. An agent writing from your brand brain has a reason to send every message: the right proof point, the right vertical, the right sender, in the tone your team already uses.
Before it writes a word, the agent reads two things. First the prospect: role, company, a recent post, a job change, the trigger that makes now the moment. Then your brand brain: which segment they fall into, the offer you lead with there, the case study and numbers cleared for that vertical, the objections that always come up, and how the named sender actually writes.
It drafts inside Sylvie: subject line options, a body that respects LinkedIn's limits, and the follow-up for when they go quiet, with each claim traced back to the deck, file or campaign it came from. Your team reviews and sends from LinkedIn as usual. What earns replies goes back into the brain, so the next batch starts from what worked instead of a blank box, and because the brain is permission-aware, one client's proof never surfaces in another client's outreach.
Each InMail opens on something true about that person: their role, a recent move, a post they wrote. No merge tag pretending to be personal.
InMail credits are finite and priced. Because every message carries a real reason to send and a matching proof point, more of the ones you spend come back as replies.
The message goes out under a real person's name, so the agent writes in that person's register. A founder's InMail reads like the founder, an account director's like the account director.
It writes for how InMail is really consumed, on a phone between meetings: a subject line that fits, a body you can scan in fifteen seconds, one clear ask.
It picks the case study, number or logo that matches the prospect's industry and company size, drawn only from what your brand brain has cleared for that use.
It writes the second and third touch too, remembers what the first one said, and changes the angle instead of resending the same pitch under 'bumping this up'.
A few of the moments where a team builds this agent instead of pasting a template into LinkedIn. Each is one-to-one outreach that would otherwise eat a senior person's morning.
The agency needs pipeline and the founder has forty target accounts saved in Sales Navigator. The agent writes each InMail from that prospect's own situation and the closest case study in the brain, so forty tailored messages go out in an afternoon and the founder spends the week taking calls instead of writing openers.
Marketing and sales agree on 25 accounts and need to reach four people inside each. The agent writes per person and per role, so the CFO gets the cost argument and the head of ops gets the workflow one. The account hears one coherent story from four angles rather than the same paragraph four times.
You have 300 seats and two weeks. The agent drafts invitations that say why this session matters to that specific guest, given their role and what they have been posting about, which lifts acceptance well past a mass invite and fills the room with the profiles sales actually wanted.
A deal went quiet four months ago and the champion has since changed jobs. The agent reads the history in the brain (what was proposed, what stalled it, what has changed since) and writes a restart that names the real reason, so the conversation reopens instead of dying in another 'just checking in'.
You are hiring three strategists and the good ones ignore recruiter spam. The agent writes from your actual culture, the work in your portfolio and the role's real scope, so the replies come from people who already understand why your shop is worth a conversation.
A B2B agency runs outbound for six clients, each with its own offer, ICP and proof. The team builds the agent once and points it at each client's brand brain, so every roster gets messages in its own voice and no client's numbers ever appear in another client's InMail.
Those write from whatever is on screen and forget it the moment you close the tab. This is an agent you build on your own brand brain, so it already knows your segments, the offer you lead with, the case studies cleared for each vertical and how your senior people write. The personalization comes from your memory, not from a prompt you retype every time.
No. Sponsored Messaging is a paid ad format that runs at campaign scale. InMail is one-to-one, sent under a named human, and each message costs a credit, so this agent is built for a different job: fewer messages, each written for a specific person. Plenty of teams build both agents on the same brand brain and let each do what it is good at.
You do not. Sylvie does not ship a LinkedIn InMail agent; this page is an example of what a marketing or new business team assembles on the platform. Sylvie wires it to your brand brain, you decide what it can read and what it can send, and the agent you end up with is yours.
Before drafting, it pulls the parts that matter for outreach: your ICP and segment definitions, positioning, pricing posture, the proof points approved per industry, the objections you hear most, and the sender's own voice. Each claim traces back to its source, and because the brain is permission-aware, a client's material stays with that client.
Messages read as spam when they have nothing specific to say. This one writes from a real detail about the recipient and a real thing you have done, keeps to your outreach rules, and stops when someone asks to be left alone. You review before anything sends: the agent drafts, your team decides.
That is what building your own agent means. You set the rules it works inside: message length, banned phrases, which case studies are usable for which vertical, how many follow-ups, which senders it writes for. As replies come in and your team edits, the brain updates, and the drafts move closer to what actually books meetings.
Book a demo and we will map the workflows worth turning into agents, and show how each one runs on your brand brain.
Request a demo