In a recent LinkedIn post, Yonathan Cohen delves into the practical application of AI sales agents, drawing insights from observations of Elon’s AI company, xAI. Cohen highlights that xAI is not just experimenting with AI but actively deploying sales agents as part of their enterprise sellers’ daily workflows. He breaks down how these AI agents can be structured into distinct roles, offering a blueprint for businesses looking to leverage similar technology.
Cohen emphasizes that while some advanced AI setups require a complex technology stack, a significant portion can operate on tools already in use by many organizations. This accessibility is a key point for businesses considering AI adoption. As Yonathan Cohen notes:
The five roles, ready to paste into whatever you already use: https://tally.so/r/J949JJ
Yonathan Cohen identifies five specific roles for these AI sales agents, aiming to enhance productivity and efficiency for enterprise sales teams.
The Chief of Staff: Orchestrating AI Efforts
According to Yonathan Cohen, the most crucial AI agent to implement first is the ‘Chief of Staff.’ This agent acts as the central hub, managing meeting preparation, inbox organization, and post-call drafting. Cohen argues that without this coordinating role, other specialized AI agents would lack direction.
→ the one she keeps pinned and uses most. It owns her meeting prep, her inbox and her post-call drafts, and it hands work to the others.→ build this one first. Without it you get nine assistants and no one running them.
In Yonathan Cohen’s view, this foundational agent is essential for the effective deployment of a broader AI sales assistant strategy.
The Prospector: Proactive Pipeline Building
Yonathan Cohen describes the ‘Prospector’ agent as a tool designed for overnight operation, ensuring a ready pipeline plan by morning. This AI agent researches potential accounts by analyzing webinars, podcasts, and social media activity, mimicking human-level research. It then drafts personalized emails in the seller’s voice.
The Account Expert: Focused Intelligence
The ‘Account Expert’ is an AI agent dedicated to specific, high-value accounts. Yonathan Cohen explains that this agent monitors relevant channels, digests calls and discussions, and provides a concise weekly rundown of new developments. The critical aspect, as Cohen points out, is its ability to report only what’s new to avoid repetition and maintain engagement.
→ that last rule is the whole trick. A digest that repeats itself gets ignored by week three.
The Forecaster: Continuous CRM Updates
Yonathan Cohen outlines the ‘Forecaster’ agent’s role in reading calls, emails, and messages to continuously update opportunity notes directly within the CRM. This addresses a common pain point where sales reps often handle such administrative tasks inefficiently, particularly at the end of the week.
→ the part of the job every rep does badly on Friday afternoon, done continuously.
The Deck Builder: Dynamic Presentation Creation
Finally, Yonathan Cohen details the ‘Deck Builder’ agent. This AI is responsible for maintaining a master presentation deck and generating customized versions tailored to individual customers and their specific language. Significantly, it can update slides in real-time from live transcripts during client calls.
→ it holds the master deck and cuts a version per customer, in their language.→ and it updates the slides while she is still on the call, from the live transcript.
Yonathan Cohen’s post provides a practical framework for integrating AI into sales operations, emphasizing role-based specialization and leveraging existing toolsets.
📝 About This Content
This article is based on insights shared by Yonathan Cohen on LinkedIn.
📅 Originally posted on September 1, 2026 | View original post on LinkedIn →