Opening Remarks

This week, I’ve been thinking about a question that doesn’t get asked enough: if emerging managers often outperform, why does most of the capital keep flowing to the same handful of funds?

My theory is that the answer has less to do with returns than most people think. Investing isn't just about maximizing performance. It's also about minimizing accountability. Backing a well-known firm is easy to defend, even when the outcome is mediocre. Backing an emerging manager is different.

That doesn't make the decision wrong to back well-known funds. In many cases, it may be entirely rational. But it does help explain why capital often concentrates even when the data suggests there are compelling opportunities elsewhere. Now, onto this week’s feature:

-Brett

Bulletpitch’s publication covers the hottest early-stage startups before being picked up by larger media outlets. If that’s you, apply here.

Setting the Scene

  • AI is quickly becoming your second brain, but the question is who owns it.

  • The AI revolution has largely been built on a tradeoff: sharing more of your data in exchange for smarter, more personalized experiences.

  • As AI becomes more integrated into daily life, users are contributing more of their emails, files, conversations, and memories to systems they don't own.

  • This week’s company is building a privacy-first local AI platform that gives users the benefits of modern AI without giving up ownership of their digital lives.

In a Sentence

Companion Intelligence is building a local AI platform that combines private infrastructure, personal AI agents, and a persistent digital memory into a single system users own and control.

  • Local AI: AI that runs on a physical box you connect to your computer that users own and control, keeping models, data, and automations on-site rather than in the cloud.

  • AI Agents: Specialized AI assistants that can perform tasks, answer questions, automate workflows, and take action on your behalf.

  • Digital Memory: A searchable, AI-powered record of your files, conversations, photos, notes, and activities that gives AI long-term context about your life.

Bulleted Version: Companion Intelligence is like having a private data center, executive assistant, and photographic memory, all powered by infrastructure that works for you instead of a technology company.

The Basics

  • Industry: Personal AI, Local AI Infrastructure

  • Headquarters: Los Angeles, CA

  • Year Founded: 2025

  • Employee Count: 12

  • Amount Raised: Bootstrapped

  • Business Model: Bundled hardware & software subscription

  • Early Traction: $500K+ ARR with accelerating sales growth

Event Board

  • Bulletpitch Cannes Lions Lunch, France: On June 23rd during Cannes Lions, we’ll be hosting an intimate omakase lunch experience on a private terrace in Cannes. This will be a highly curated, invite-only gathering of top creators, founders, and executives. Minimal spots remain. Reply to this email if interested in attending.

Weekly Feature Continued

Due Diligence

WHAT WE LIKE
  • Large Emerging Market: Personal AI is evolving into a new computing category, combining personal computing ($155.19B) and personal AI assistants ($14.67B), creating a large opportunity for platforms that own the relationship between users, data, and AI.

  • Growing Trends: Companion Intelligence is positioned at the intersection of three major trends: AI adoption, growing demand for data ownership, and the shift toward privacy.

  • Integrated Product Strategy: By controlling the hardware, operating system, agent ecosystem, and data layer, Companion Intelligence can deliver an experience that software-only competitors may struggle to replicate.

POTENTIAL RISKS
  • Consumer Education: Most consumers have never considered owning their AI infrastructure, requiring Companion Intelligence to not only sell a product but help define an entirely new category of personal AI ownership.

  • Hardware Complexity: Building and supporting a physical appliance adds capital constraints, inventory, manufacturing, customer support, and upgrade-cycle challenges that pure software AI companies do not face.

  • Energy Usage: Running a local AI appliance could create meaningful electricity costs, especially for users operating larger models, multiple agents, or always-on automations.

Founder Profile

To request an introduction to the founder, respond to this email.

Comps

  • Rabbit: Consumer AI hardware for interacting with apps and digital services; Companion Intelligence differs by using local compute and private memory for longer running agentic tasks.

  • Sanctum: Software that lets users run open-source AI models locally on their own devices, less focused on custom agentic software and dedicated hardware.

  • Limitless: Consumer AI wearables that capture and summarize conversations as a personal memory layer; potentially a valuable data source Companion Intelligence could integrate into its memory layer.

Why Companion Intelligence: By combining local infrastructure, AI agents, and Digital Memory into a single platform, Companion Intelligence is thinking outside of the box to solve personal AI ownership.

Cast Your Vote

What do you think of Companion Intelligence?

Cast your vote below and tell us why:

Login or Subscribe to participate

Last Week Today

  • The Results Are In: PLANET.TECH, a company building the operating system for cattle by turning methane emissions into renewable fuel revenue and using real-time health intelligence to help farmers protect their herds, was contested in last week's poll.

  • Subscriber Feedback: “Unproven hardware adoption among farmers is an extremely steep hill to climb.”

Keep Reading