Want to connect with Erdos?
Join organizations building the agentic web. Get introductions, share updates, and shape the future of .agent.
Is this your company?
Claim this profile to update your info, add products, and connect with the community.
Erdos represents a specific application of AI agents in the domain of real-time performance coaching. In the AI agent stack, Erdos acts as a specialized analytical agent that perceives a stream of high-speed environmental data (gameplay telemetry), processes it against strategic models, and suggests optimal actions or refinements. It is part of the broader trend of "copilots" or agents that assist human performance in digital environments.
For the agent ecosystem, Erdos is a useful case study in how agents can be deployed in environments with low latency requirements and complex, multi-dimensional data. While it is currently focused on the gaming vertical, the underlying architecture of processing real-time events to provide prescriptive feedback mirrors the needs of agents in professional and industrial settings. They are championing the transition from static dashboards to active, agentic advisors.
Gaming data has historically been a graveyard of spreadsheets and raw numbers. While platforms like Overwolf and various API-driven stat trackers have existed for years, they typically offer descriptive analytics; they tell a player what happened without explaining the tactical adjustments required to improve. Erdos, the flagship product of Shapley Technologies, is an attempt to move the category into prescriptive analytics. The company builds a platform designed to interpret gameplay data in real time and provide specific, actionable insights for competitive gamers.
The platform functions by ingesting match data and player performance metrics to identify patterns that human coaches or the players themselves might overlook. Rather than simply reporting a win-loss ratio or a kill-death spread, the Erdos AI analyzes the mechanics behind those numbers. This includes strategic decision-making and tactical positioning. The goal is to democratize high-level coaching, which has traditionally been a resource reserved for professional esports organizations. By automating the analysis of gameplay, the company targets a broad user base ranging from casual players looking to improve their rank to semi-professional competitors.
Shapley Technologies was founded in 2023 by Jonathan Sirotin and Angelica Sirotin. Jonathan Sirotin is the CEO and brings a background that is deeply rooted in the esports ecosystem. Before founding the company, he was a professional coach and manager, notably scaling Alpine Esports into a top-ten team in North America. His experience as a semi-professional Rocket League player and tournament analyst informs the product's focus on practical, coach-like feedback. Angelica Sirotin, the co-founder, comes from a background in international business analysis and venture building, specifically focusing on AI investment through her firm, Sirotin Ventures.
The company’s trajectory is influenced by a board that includes Charlie Jackson. Jackson is a veteran of the software industry, having founded Silicon Beach Software and co-founded FutureWave Software. The latter is significant in tech history as the creator of the technology that eventually became Adobe Flash. This connection suggests a strategy focused on building foundational software tools rather than a narrow gaming utility.
In the competitive market, Erdos sits between general-purpose gaming trackers and high-end professional coaching services. It competes indirectly with established entities like Tracker Network or Mobalytics, which offer statistical databases but often lack the intelligent layer that Erdos aims to provide. The challenge for the company lies in the technical integration across various game titles, each with its own data constraints and API limitations. Currently, the company focuses on delivering personalized insights that help gamers refine strategies and optimize their performance through an automated, AI-driven interface. Based in the United States and backed by experience in both the competitive gaming scene and the broader software market, Erdos is positioning itself as a tool for the next wave of intelligent gaming assistants. It moves the needle from simple data reporting to active performance guidance.
An AI-powered performance platform for competitive gamers that provides instant, data-driven insights.
A collection of optimization problems in mathematics
A community database for the problems on the erdosproblems.com site
A Lean companion to Analysis I
Code to automatically prove or verify estimates in analysis
Project to verify the Guy-Selfridge conjectures
A project to map out the relations between different equational theories of Magmas.
Exponent pair database
Erdos is hiring
You've explored Erdos.
Join organizations building the agentic web.