# Knowledge Systems: The New GTM Stack — Jeffrey Wang, Exa

## Executive summary

Jeffrey Wang argues that Go-To-Market (GTM) strategy must be treated as an AI engineering problem. The core thesis is that GTM is fundamentally a data problem, requiring the creation of a 'live model of your world' that autonomous agents can act upon. He details systems like Exa (a search engine for agents), the ICP dashboard for classifying the Total Addressable Market (TAM), and Request Lens for real-time customer signal detection. Key architectural principles include making the system API-first, recognizing that consistent UIs still complement flexible chatbots, and prioritizing arbitrary customizability over rigid build vs. buy decisions.

## Key takeaways

- GTM as a Data Problem: The goal is to build a live model of the world—combining internal data (customer usage) with external data (web activity, company information)—that agents can programmatically act on. This shifts GTM from a purely sales function to an engineering challenge.
- Agent-First Requires API-First: For any agent system (whether it's a GUI or a chatbot) to access data, the underlying systems must expose robust programmatic interfaces (APIs). This is critical for enabling agents to function.
- System Components: ICP Dashboard & Request Lens: The ICP dashboard uses Exa's embeddings over the internet to classify every company in the TAM and estimate anticipated spend. Request Lens provides real-time alerts when significant customer signals occur (e.g., signups, search surges).
- The Value of AI Cloning (Jeffbot): An agent can be trained on historical data to mimic a user's professional style and decision-making. Jeffbot was built by analyzing 760 emails and hundreds of past decisions, creating 'evals' to calibrate its judgment against the founder’s own behavior.

## Technical details

- Exa Architecture: Exa is described as a 'search engine for agents.' Technically, it functions by taking the internet and training embeddings to provide arbitrarily powerful semantic filtering and slicing of data.
- Agent Capabilities: Agents can be used across various workflows, such as account executives building demos or drafting communications (e.g., Jeffbot drafting Slack messages). The system must manage security boundaries; for instance, Jeffbot has full read/write access when called by the founder, but others are limited to drafting.
- Decision Modeling: To create a sophisticated agent clone (like Jeffbot), one must analyze historical artifacts—specifically decisions made in Slack and email—and generate 'evals' to calibrate the agent’s judgment.

## Practical implications

- Treat GTM strategy as a data engineering problem by building a comprehensive, live model of the entire addressable market (TAM).
- Prioritize API development across all internal and external systems to ensure agents have programmatic access to necessary data.
- Design workflows that integrate AI agents into existing operational tools (e.g., Slack, CRM) rather than relying solely on standalone chatbots.
- When selecting enterprise software, focus on platforms that offer arbitrary customizability via exposed APIs (MCPs) rather than assuming a single 'build vs. buy' choice.

## Topics

AI Engineering, Data Modeling, Agent Systems, Go-To-Market Strategy, Information Retrieval, Exa, Cursor, Cognition

Source: https://www.youtube.com/watch?v=6pbQgnJ9Voc