What does SDR stand for?
Sales development representative: the early-sales role that finds prospects, starts conversations, and hands interested buyers on. An AI SDR is software that does that work.
AI sales development representative
What is an AI SDR?
An AI SDR, short for AI sales development representative, is software that does the early-sales work a human SDR does: it finds prospects, researches them, writes first-touch outreach, follows up, and sorts the replies so a person can take the conversations worth having. A good one keeps a human in charge of what gets sent. Zevari runs it as a full outbound department for LinkedIn and email: six specialists, and sends wait for your approval by default.
The work
What an AI SDR does.
Five jobs, the same five a human SDR does. The useful question for each is where a person stays involved.
| Job | What the software does | Where you stay involved (Zevari) |
|---|---|---|
| Find prospects | Builds lists from saved searches, filters, and buying signals such as a post asking for a recommendation. | You say who you sell to. |
| Research them | Reads the profile, the company, and recent activity before anything is written, then scores fit against your ideal customer. | You correct the scoring when it is wrong, and it learns. |
| Write outreach | Drafts a first touch from that context, in your voice rather than a template. | You approve or edit each draft before it sends. |
| Follow up | Runs the next steps of the sequence on a sensible cadence and stops when someone replies. | You approve each follow-up too. |
| Sort the replies | Reads every reply and flags genuine buying interest apart from auto-replies, pitches, and polite no's. | You take the conversation from there. |
The category
AI SDR, AI sales team, sequencer, LinkedIn MCP.
Four terms that get used as if they meant the same thing. They do not.
- Sequencer
- Sends a fixed sequence to a list. It automates the sending, not the judgement: no research, no fit scoring, no reading of replies.
- AI SDR
- Software doing the SDR jobs end to end. Quality varies most in two places: how much of its work you see before it sends, and what it counts as a result.
- AI sales team
- The SDR job split across specialists, each owning one step and handing off to the next, so you can see where a prospect is and why. This is how Zevari is built.
- LinkedIn MCP
- A connector that lets an AI client such as Claude, Codex, or ChatGPT call LinkedIn actions. It is a way to drive a team, not a team.
More terms, defined, in the glossary.
What a department adds
One AI SDR, split into six specialists.
A full outbound department, running today. Zevari gives each SDR job to a specialist that owns one step and hands off to the next, on LinkedIn and email, so you can see where every prospect is and why.
01 · Prospecting
Fills the list while you sleep.
Runs saved Sales Navigator searches and lands net-new people on your lists.
02 · Signals
Finds the reasons to reach out.
Finds LinkedIn posts by keyword and captures people who engage with your posts.
03 · ICP Scoring
Sorted by what actually closed.
Scores every new lead against your ICP and learns from your reviews.
04 · Voice Writer
Writes messages that land.
Drafts personalized, on-brand outreach for every step.
05 · Follow-up
Keeps it moving.
Runs your campaign sequences and shows what needs your attention.
06 · Reply Qualifier
Finds the conversations that matter.
Assesses replies automatically and highlights buying interest.
Boundaries
Where an AI SDR should stop.
It should not send unsupervised.
Unattended sending is how LinkedIn accounts get restricted and how bad messages go out under your name. At Zevari, every connection request, message, comment, and post waits for your approval by default. No specialist moves to autopilot until it has earned it, and you can take it back to reviews with you at any time.
It should not promise meetings it cannot show you.
Zevari counts qualified conversations, under a published, versioned definition: the buyer replied with genuine interest and kept talking, and the conversation traces to outreach Zevari sent. Not every reply, and not a booked meeting.
It should not put the account at risk.
Zevari is a hosted connector: no browser extension and no cookie injection. Limits are enforced in code, not left to a slider: 150 connection requests a week on a Free* or Premium account and 200 on Sales Navigator, a daily cap on all actions of 400, 1,000, or 2,000, burst limits per minute and per hour, and activity inside your working hours. The weekly figure covers connection requests sent without a note: LinkedIn caps requests that carry a note at 5 a month on a free account, and gives Premium and Sales Navigator a far larger allowance, so Premium or Sales Navigator is the practical choice for real outbound. The safety page has the full table.
Buying one
Six questions to ask any AI SDR.
- Who approves what gets sent, and can anything go out without that approval?
- How does it connect to LinkedIn? A browser extension or injected cookies put the account at more risk than a hosted, session-based connection.
- What limits does it enforce in code, per day and per week, and at what numbers?
- What does it report as a result? A reply, a meeting, and a qualified conversation are different things, and a vendor should say which one it counts.
- Does it write in your voice, or from a template library other companies are also sending?
- Can your own tools reach it, over MCP or an API, if you want to drive it yourself?
Price
$87 a month, or $497 a year.
One plan with all six specialists and one LinkedIn account. The yearly rate saves 52%. Extra LinkedIn accounts are $37 a month. 7-day free trial for first-time accounts, 30-day money-back guarantee after the first charge.
See pricingRun it yourself
Drive the same team from your own AI.
Claude Code, Codex, ChatGPT, or any MCP client can drive the six specialists through the Zevari LinkedIn MCP; your own code can use the REST API. Same plan, same approvals.
Limitations
What this page does not claim.
- No reply-rate or conversation figures are published here yet. When Zevari publishes one, it will name the population, the timeframe, and the definition version it uses.
- Fit scoring is an assessment against the ideal customer you describe, not a fact about the person. You can correct it, and it learns from your corrections.
- A qualified conversation is an early-funnel outcome. It is not revenue, and it is not a guarantee of a meeting.
Questions
AI SDR FAQ.
Is an AI SDR the same as an AI sales agent?
Mostly, yes. The terms are used interchangeably. What matters more than the label is what it sends without you, how it connects to your accounts, and what it counts as a result.
Does Zevari's AI SDR send messages on its own?
Not by default. Every connection request, message, comment, and post is staged for your approval before it reaches your account. A specialist moves to autopilot only after it has earned it, for Follow-up ten approved sends in 30 days, and you can switch it back to reviews with you at any time.
Does Zevari book meetings?
Zevari reports qualified conversations: a replying buyer with genuine interest who kept talking, traced to outreach Zevari sent. That is a published, versioned definition. It is not a booked meeting, and Zevari does not use the two terms interchangeably.
Does it work on email as well as LinkedIn?
Yes. LinkedIn is the primary channel. Add a mailbox and the same team runs email steps in the same sequence, under the same approvals and daily limits.
How much does Zevari's AI SDR team cost?
$87 a month, or $497 a year (save 52%). All six specialists and one LinkedIn account are included; extra LinkedIn accounts are $37 a month. First-time accounts get a 7-day free trial, and there is a 30-day money-back guarantee after the first charge.
Can I run it from Claude Code or Codex?
Yes. Technical teams drive the same six specialists from Claude Code, Codex, ChatGPT, or any MCP client through the Zevari LinkedIn MCP, or from their own code over the REST API. Same plan, same approvals.