CORE GTM brain
CORE is a self-hostable personal AI assistant for technical users that remembers context across apps and executes delegated tasks.
category
Personal AI assistantsmarket profile
- text: AI users juggle context across email, meetings, development tools and assistant conversations.
- text: Personal agents compete on persistent memory, app access, task execution and control of deployment.
personas
- title: Developer and technical founderpains:
- Repeatedly reconstructing project context for coding agents
- Small issues and pull requests compete with focused work
- title: AI power userpains:
- Assistant conversations lack shared memory across tools
- Switching between messages and apps fragments tasks
- title: Privacy-conscious self-hosting operatorpains:
- Personal context and credentials span many third-party tools
- Wants control over where assistant data is stored
icp
- text: Individual developers and small software teams using GitHub, issue trackers and coding assistants daily.
- text: Technical users able to run Docker or provision a server and willing to connect their apps.
message pillars
- text: Keep assistant context across tasks.
- text: Retain control through self-hosting.
- text: Delegate work to coding agents.
- text: Reach the assistant through familiar channels.
objections
- objection: Will setup be too technical?response: The repository provides a Docker setup wizard and Railway deployment path; validate infrastructure requirements first.
- objection: Can I retain control of sensitive context?response: Self-hosting keeps data in your infrastructure; review permissions and the security policy before connecting accounts.
- objection: Will it act without oversight?response: The documented task flow includes plan approval before execution.
competitors
- name: OpenClawwhy: Open source personal agent with terminal and browser access.
- name: Hermes Agentwhy: Open agent offering persistent memory and assistant workflows.
- name: Devinwhy: Delegated coding agent competes for software task execution.
keywords
- self hosted AI assistant
- personal AI memory
- open source AI agent
- AI assistant app integrations
- AI coding agent task automation
- persistent memory for AI
voice
- Personal and ownership focused
- Technical and builder oriented
- Aspirational about proactive help
signals
- text: Developers already use multiple coding agents and complain about repeatedly supplying context.
- text: A technical user is looking to self-host a personal assistant that connects work applications.
stack
- Docker
- GitHub
- Linear
- Slack
- Claude Code
- Codex
- MCP
pricing
model: unknown
motion
community_ledchannels
- GitHub
- Discord
- developer community
confidence
mediumFounders
- name: Manik Aggarwaltitle: Founderrole: Founderbio: Manik was a Senior Product Manager at Velocity. He scaled Velocity Insights from 0 to 1500 companies. Prior to joining Velocity, Manik worked as a Data Scientist at American Express. During his tenure, he built and maintained global credit risk models.
- name: Manoj Konchatitle: Founderrole: Founderbio: Manoj is an Engineering Leader with 7+ years of experience, successfully led data engineering and backend teams at Velocity and Emtropy Labs. At Velocity, he built robust data pipelines, effortlessly handling millions of rows daily basis. He engineered an NLP model that drastically reduced recruitment time.