One question first. Grouped, and answered short.
Replicate Labs is the GTM harness. It gives the frontier models your buyers, your method and your scorecard, then coaches reps by voice on live deals, writes the follow-up, and reads the business for the people who run it. Five groups below; open the one that has your question.
The product.
What it is, what it sits beside, and what it does that the tools you already run do not.
What is Replicate Labs?
Replicate Labs is the GTM harness for enterprise revenue teams. A harness is the runtime around a model: it decides what the model is fed, which tools it can call, what happens to the output, and what happens when the output is wrong. We claim no model of our own. We claim the harness around the frontier models that makes them good at your selling. It installs your buyers, your method and your standard once, then coaches reps and managers by voice on live deals, writes the follow-up and the proposal, and reads the business for the people who run it. The named coaches, Keenan, Val, Caty and Bryce, sit inside a rep’s week: before the call, on the live deal, after the call. The job is to close the gap between what a rep learned in training and what they do on a real call.
Is it a training tool or a coaching tool?
Both, and that is the point. The harness runs across the rep’s journey: building the skill in a roleplay, embedding the behaviour in a course, and coaching execution on live deals. Those layers share context. What a coach learns about a rep in a practice session informs how it coaches them on a real deal weeks later.
How is this different from an LMS like Mindtickle or Showpad?
An LMS measures whether a rep watched something: modules, quizzes, certifications, completion rates. The harness measures whether a rep can execute, in a roleplay scored against your scorecard and on the live deal. The distance between passing a quiz and winning a deal is the part it closes. Content delivery gets you compliance. This gets you execution.
How does Replicate Labs compare to Gong?
Gong is conversation intelligence. It records, transcribes, and tells you what happened. The harness takes what happened and turns it into a change in the next call, an email judged before it sends, and a read for the manager. They work well together. Gong recordings flow into the harness as deal context, so the coach knows what the buyer said on the call.
Does this replace my sales manager?
No. It takes the parts of coaching a manager cannot physically scale. A manager with 15 reps cannot run a coaching session before every important call. A coach can. That gives the human manager room for what only a person can do: strategy, deal escalations, and the conversations that need one.
How does the AI know about my deals?
Through the context you connect. CRM integrations (Salesforce, HubSpot) sync opportunity stage, value, contacts and history. Call recorders (Gong, Fathom, Fireflies, Salesloft) bring in what was said on the call. And every coaching conversation adds to a growing deal record. By the time a rep asks for help, the coach already knows the deal.
What does it integrate with?
CRM: Salesforce and HubSpot. Call recording: Gong, Fathom, Fireflies and Salesloft. Calendar: Google and Microsoft. And Slack, so coaching happens where the rep already works. You can talk to a coach by video, by voice, or by typing when you cannot talk.
What an answer looks like. A question about the evidence, then one move.
Ask a coach about a deal and the first thing back is a question about what the buyer said. The deal review itself is a call, transcript open; typing is the fallback for when you cannot talk.
Call the coach with the deal. Before the meeting and after it.
Roleplay the next conversation. Against a buyer built from your personas, scored on your scorecard.
The follow-up, judged before it sends. In the buyer's words, in your method.
Your coachLive deal · CorranNothing yet. Find the sentence first.
When Tessa said this mattered, what did she say, in her words? If that line is not in a transcript or an email, you are forecasting on your own read. Pull the 2 September call, find where she put a cost on carrying on as they are, and bring that line to a call.
[Corran call, 2 Sep] [Email thread, 5 Sep]The coaches.
Four named coaches, each carrying one method and one job. A deal review is a call with one of them.
Which coach should I talk to?
Keenan if you sell with Gap Selling. Caty if your team runs MEDDPICC, MEDDIC, SPIN, or a custom playbook. Val if your team runs the ValueSelling Framework®. Bryce if you are a sales leader who wants pipeline and team-pattern analysis. The AI Managers page walks through the choice in detail.
What does an answer from a coach look like?
A question about the buyer’s evidence first, then one move. Ask a coach what to send a champion who has gone quiet and it asks what the buyer said when they told you this mattered, in their words. If that line is not in a transcript or an email, you are forecasting on your own read, and the coach says so. Then one thing to do before the next conversation. The deal review itself is a call, with the transcript open; typing is the fallback for when you cannot talk. The scene above this group shows one.
Can I talk to a coach outside the platform?
Yes. The same coach, with the same context, answers in Claude, in ChatGPT, in Slack and from the CRM. A manager asking Bryce in Slack pays nothing extra on a Platform seat. Voice, video, roleplay and Dynamo stay in the platform.
The method.
The selling approach your team runs is the one the coach applies.
What sales methodologies do you support?
Gap Selling, coached by Keenan. The ValueSelling Framework®, coached by Val. MEDDPICC, MEDDIC, SPIN, CHAMP and custom playbooks, coached by Caty. On the outbound side, Gap Prospecting pairs with Keenan and Vortex Prospecting pairs with Val. Each coach is built to hold its methodology in every conversation.
Can I use two methodologies at once?
Each coach has one core methodology. Keenan is built on Gap Selling, and if you ask he will also coach against MEDDPICC, MEDDIC or SPIN alongside Gap, never instead of it. Caty is methodology-flexible and configured to your playbook. If you want a coach whose primary methodology is one of those frameworks, that is Caty or Val.
What is the standard a call is scored against?
Your scorecard. Without one, a model has nothing to judge work against and agrees with the rep. The harness installs your scorecard once, then every roleplay round, every judged email and every capability read is scored on it. Where a team has no scorecard, the methodology partner’s is the starting point.
The data.
Where your deals and calls go, who can see them, and what they are never used for. The long version is the security page.
Security and trustIs my data used to train AI models?
No. Customer data is not used to fine-tune or train any language model, by us or by the providers that answer the coaches, whose enterprise terms commit to the same. All personalisation happens at inference time, through the harness and a per-user memory. Your team’s data does not feed a shared model that other customers benefit from. Every provider that touches your data is named on the data page and in Annex 2 of the DPA.
Is Replicate Labs secure?
Replicate Labs holds a SOC 2 Type 2 report, examined by INTERCERT CPA LLC for the period 2 March to 8 July 2026, clean with no exceptions raised, and is UK GDPR compliant and headquartered in London. A penetration test by Astra Security in July 2026 graded the platform A+. Data is encrypted at rest and in transit, infrastructure runs on Google Cloud with a London region by default and a US region on request, and accounts are isolated at the company level with no cross-customer access through the application. Vulnerability scanning runs at least monthly and penetration testing annually. The security page carries the detail and the review pack is available under NDA.
Who in my company can see my coaching?
It depends on the account. On Platform, the account owner gets Bryce and team analytics, so they can see coaching activity across the seats they pay for. No user or coach in any other account sees it. The data page sets this out by account.
The money.
Two ways to buy.
What does it cost?
Replicate Labs is bought two ways. Platform is seats, sold with implementation, set up with your methodology partner and quoted on the team, the method and the integrations. Builder is pre-purchased credits for running the harness as infrastructure in your own systems: the outbound agent you build on the coach’s judgement, a scored call archive, the asset after every demo. Neither has a public price. The pricing page sets out what each includes.
How do I get started?
Bring a real stalled deal to a call with us and watch a coach work it while you listen. That is how Platform starts, with Keenan, Val or Caty.
Still a question? Ask it on a call, with the deal that prompted it.
Your pipeline, the method your team already runs, and a coach working a real situation while you watch. You leave with the answer and something to evaluate.