Hey, it's Matt here. Really excited to launch Mocke today, because it has been a wow moment for ourselves and beta users. Can't wait to share it with everyone who writes cold email.

TLDR;

Mocke is an AI agent for cold email. It works like Cursor for development. Mocke collapses hours of cold email writing into seconds and conducts deep research on each lead to deliver completely personalized messaging at scale.

Try Mocke here https://mocke.co/. No account is needed; just enter your website to see the cold email you are going to love in 60 seconds.

The problems we are solving

The 0.1 → 1 Problem

Even though everyone now has access to ChatGPT, crafting effective cold emails still requires significant time and effort. Here's the typical experience of someone writing cold emails:

1. ChatGPT provides a first draft in 5 seconds, which looks decent.
2. The user then spends time refining the phrases, CTA, subject line, messaging, and final touches. This process can take an hour or even several hours until they feel comfortable sending out the emails.

The challenge is rather the effort needed to refine the draft for production, because no one feels comfortable taking the 1st drafted email from ChatGPT and send to real leads. Therefore, I refer to this problem as the 0.1 to 1 problem.

The personalization with deep research

ChatGPT is a commodity that everyone already has access to. So, writing a good cold email is no longer sufficient to get meaningful returns from campaigns. Relevance has already become the #1 indicator for a high reply rate. However, there are no effective tools available for conducting in-depth research on each lead and crafting personalized, unique emails for every individual.

In the market, many AI sales tools attempt to build rapport in cold emails by referencing LinkedIn activities, such as congratulating someone on a promotion or mentioning a recent post. However, leads can quickly recognize poorly generated personalization in these outreach efforts, which often lack genuine relevance. This undermines trust and reduces interest.

The solution - Mocke: your cold email AI agent

Mocke is a Cold Email AI agent that functions like Cursor. It reduces hours of cold email writing and deep personalization to just 60 seconds, delivering thousands of unique email messages that leads will love instantly. While it works alongside you, the AI agent handles most of the ideation, writing, polishing, personalization, and deep research, allowing you to achieve results without lifting a finger.

Step 1: Campaign idea

Mocke suggests campaign ideas. By simply entering your website, the AI agent identifies the ideal ICP, highlights the most suitable pain points for cold emailing (as not every pain point is effective for cold emailing), and provides a creative messaging angle designed to attract a lot of responses.

You can start not only with your website but also with your inbound website visitors, conference pre- and post-outreach, publishing a new case study, creating a new blog post, promoting a new AI feature, and more. Share any of these with Mocke, and the AI agent will develop a comprehensive outbound strategy for you.

Below is a basic example using @Pylon’s website. (big fan of their product btw!)

Step 2: Email template

Mocke writes email templates with embedded prompts. Refer to the Pylon example below.

For each email, AI agent may propose an engaging subject line, a creative value-based call to action (beyond a simple meeting request), and crucially, instructions for conducting deep research and personalization, presented as prompts in [square brackets].

What Mocke will do is to run deep research into a lead, the business behind, their department and teams, and all relevant information with regard to what AI agent is selling. Then with the targeted deep research, AI agent will generate a highly-relevant email for one lead.

Just to highlight, in the Pylon example above, there are various prompts:

• A paragraph prompt [Write a short paragraph in 30 words or less. Ask the lead if their support team has experienced a large volume of support tickets in Q1 2026, and how they handle it. Must mention one specific technical product error message often asked by the lead's users. Must use deep technical terminology in the context of the lead's product. Must mention the lead's product name.]
• A phrase prompt [describe the lead's team, in 3 words or less]

So you can choose from full delegation to AI to write content, or partially if you want to have finer control over customization.

Step 3: Preview deep research and emails

To review the deep research and email writing, you can provide a lead's LinkedIn profile for the AI agent to conduct in-depth research and craft a personalized email in real-time. This ensures you feel confident about the dynamic content generated by the AI, based on the prompts embedded in the email templates.

For this demo, I'm again using two potential leads of Pylon and have AI run deep research to write emails. You can compare them side-to-side to see the difference.

Ready to launch

After completing the three steps in just 60 seconds, you can upload your list of leads. The AI agent will then conduct deep research and create personalized messages for each lead, one by one!

🚀🚀🚀 Ask

No sign-up is required, and it's free to use. Simply visit our website https://mocke.co/ and enter your website. In just 60 seconds, you'll receive a cold email that you'll truly like.

My ask: try it out with your website and share your honest feedback. Let us know whether you love it or not.

— — — —

How it works underneath

If you are still reading and interested to know a little more.

Email Writing Workflow

For each lead, Mocke would go through a few steps to write unique messaging per person and ensure it's relevant.

1. Build a profile of the lead by scraping LinkedIn page
2. Build a profile of the company of the lead by querying B2B databases
3. Use @Firecrawl (shout out to the team, amazing product as well!) to crawl website and/or targeted information search using its agent
4. Use agentic loop to finish email writing, polishing, reviewing, scoring, and delivering

What can be found in deep research?

1. Anything mentioned on main marketing website
2. Anything mentioned in documentation
3. Google searchable content if the results come back in the first page of SERP
4. Top reddit relevant threads (same reason as above, indexed by Google)
5. Personal and company LinkedIn profiles activities

What if AI can't find information?

Mocke AI agent will not generate inaccurate information, as we have safeguards in place. In the case of information not found, AI will write generic content, rather than specific reference of information.