The agents

Your agents are digital colleagues. Each one is a specialist that does a single job well - qualifying an account, researching a contact, drafting outreach grounded in real findings, or running a call drill with you before you pick up the phone. You bring the judgment and the relationships. They do the heavy, repetitive groundwork that used to eat your week.

This page is the catalog. It groups every agent by what it does and links you to the page that explains how to get the most from it. Agents augment skilled salespeople - they never replace them.

How to think about the catalog

The agents sit in three groups, mapped to the three things a sales team has to do well: find and understand the right accounts and people, reach them in a way that lands, and keep getting sharper at the conversation itself.

Research - works out who is worth your time and why now. Qualification, deep account and contact research, finding the right people, and verifying their contact details.
Personalization - turns what research found into outreach and account strategy that reads like a colleague who did their homework, not a mail-merge.
Coaching - lets you rehearse the call and get structured feedback before it counts.

Each agent draws from the same shared intelligence you set up once in the Agent Training Center - your verticals, personas, and value proposition. That shared context is why the outreach an agent writes is context-driven, not data-driven: it is built from what was actually found about that company and that person.

The agents at a glance

You pay per task, in credits, not per seat. A “task” is one agent doing its job once - qualifying one account, researching one contact, drafting one play. What a single task costs depends on how hard you ask the agent to work and how fast you need it back; see Effort and speed for that trade-off and Credits for how credits, plans, and balances work.

Research

These agents work out who is worth your time and why now.

Account Qualification - decides whether an account fits your ICP, and shows the evidence behind the verdict.
Account Research - builds a deep dossier on a qualified account: financials, tech, hiring, news, and buying signals.
Domain Finder - finds and verifies the right company website so the other agents have something solid to research.
Contact Finder - surfaces the people at an account who match your buyer personas.
Contact Qualification - confirms a contact still fits the persona and still works there.
Contact Research - profiles a person: communication style, recent activity, and conversation starters.
Email and phone waterfall - cascades through many data vendors to find a verified email or a direct dial, and only charges on a successful result.

Personalization

These agents turn what research found into outreach and account strategy.

Play Copywriting - writes the email, call script, or message from what research actually found.
Account Planning - builds a strategic account plan: stakeholder map, competitive read, and approach.
Digital Twin - a conversational colleague you chat with to prep, handle objections, and sharpen positioning before a call.

Coaching

This agent area lets you rehearse and get feedback before the conversation counts.

Coaching: roleplay and feedback - practice cold calls and discovery against a realistic buyer, then get scored, structured feedback against your own playbook.

How you actually run them

You almost never run agents one at a time by hand. Most teams chain them into playbooks - pre-built sequences that handle a whole use case end to end, like qualifying and researching a list of event attendees, or watching for buying signals and acting the moment one fires. Reps also run individual agents on demand while they work, including from the Chrome extension right alongside a prospect’s profile.

How thorough and how fast an agent works is something you control per run. You set an effort level - the deeper the reasoning, the more it costs - and a speed, where a faster turnaround applies a higher cost. See Effort and speed for the trade-off between depth, turnaround, and credit cost, and the detailed effort and speed cost table for the exact per-run numbers.

Eva: useful framing for a sales reader - research agents answer “who and why now,” personalization agents answer “what do I say,” and coaching agents answer “am I ready to say it.” If a customer asks which agent to start with, the honest answer is usually “you start with a playbook, not an agent,” because playbooks orchestrate the right agents in the right order. Point them to the playbooks page.

On per-agent credit cost, do not quote a single fixed number per agent in the catalog. The cost of a run is the effort credit cost times the speed multiplier (see the effort and speed cost table). A few agents expose effort directly with their own credit cost per level - for example Domain Finder is 1 credit at low effort and 2 at high effort, and Contact Finder is 1, 2, or 3 credits depending on effort. The waterfall is the exception to the effort model: it charges only when it returns a verified result, and the email and phone lookups are billed separately (its own page carries the per-lookup detail). If a customer needs an exact figure for another agent, route them to the effort and speed cost table rather than guessing.

In practice

A rep working a new account rarely thinks in agents. They run a playbook against a list, and behind the scenes the right agents fire in order: the account gets qualified, the qualified ones get researched, the right people get found and verified, and a first draft of outreach lands - already grounded in a funding round, a leadership change, or a recent post, not a generic merge field with a blank name in it. The rep reviews, adjusts, and sends with confidence. Every agent in that chain is one of the colleagues catalogued above.

To see how these pieces stack on top of each other - account fit, then signals, then people, then outreach - read How the layers fit together.

Running agents - how to kick off an agent on demand, including from the browser extension.
Effort and speed - the depth-versus-turnaround choice that sets each run’s cost.
Effort and speed cost table - the exact per-run credit figures for every effort and speed combination.
Credits - how credits, plans, and balances work across all agents.
Playbooks - the pre-built sequences that orchestrate these agents end to end.
Agent Training Center - the shared context every agent draws on.
Context over data - why agent output is grounded in real findings, not merge tokens.