Grid Heap · An AI agent company
Agent infrastructure.
Agent applications.
Grid Heap builds the LLM gateway, execution layer, and omnichannel messaging layer that production AI agents run on, and builds vertical and consumer agents on top of them. The same APIs are open to developers.
Agent infrastructure
The model gateway, execution layer, and messaging layer behind our agents. The same APIs are open to developers.
AIRuntime
LLM gatewayAccess major AI models through one OpenAI-compatible API, with routing, failover, usage metering, and cost controls.Explore AIRuntimeAgentMQ
Execution layerRun agent workflows under explicit permissions and budgets, with approvals, idempotent tools, human checkpoints, and audit trails.Explore AgentMQTidyMessages
Messaging layerUnify iMessage, WhatsApp Business, and web chat, with shared conversation state, webhooks, and channel compliance.Explore TidyMessagesVertical agents for business
Built and run on our own infrastructure.
TidyAnswer
AI front deskAnswer customer messages from your knowledge base, across channels. Hand off to your team with full context.Explore TidyAnswerRShift
Advertising & brand marketingResearch markets, create content, and run cross-platform ad campaigns in one agent-powered workspace.Explore RShiftSpeech Name
Voice marketplaceSell your voice for a revenue share. Voice-cloning APIs watermark every output to verify authorized use.Explore Speech NameConsumer agents
Agents for personal knowledge, conversations, planning, and the work ahead.
nnu.us
Agent-powered bookmarksSave a link. Agents crawl, summarize, and translate it for you.Explore nnu.usEnola
Personal AI assistantYour personal AI assistant on iMessage and WhatsApp.Explore EnolaSuperPlanner
Chat & planningAgent chat and planning in a notes interface.Explore SuperPlannerTrainShell
Agent-integrated terminalTrain models and develop ML projects with agents in your terminal.Explore TrainShellDueglow
Countdown productivityStay focused with countdowns for your plans and deadlines.Explore DueglowResources
Useful data and creative assets for founders, builders, and designers.
heap.page
Deals & perks for foundersFounder deals and perks, sourced, verified, and updated by agents.heap.page, opens in a new taboutseek.lol
Product data & rankingsLive product data and rankings, continuously tracked by agents.outseek.lol, opens in a new tabIcons
Icon libraryFind, customize, and export icons with source and license information.icons.2d.digital, opens in a new tabEffects
Image effectsApply visual effects to images and logos, right in your browser.effects.2d.digital, opens in a new tabSounds
Sound libraryFind interface sounds in a searchable, CC0-first library.sounds.2d.digital, opens in a new tabModels are ready.
Production is the challenge.
Model capability is no longer the constraint on AI agents; production readiness is. An agent in production needs provider-agnostic model access with cost control, an execution layer that runs tool calls under explicit authorization with an audit trail, and delivery on the messaging channels people already use.
Grid Heap builds all three, because our own agents needed them first. Every application runs on the same infrastructure in production, and that infrastructure is exposed to third-party developers through the same APIs.
- Our mission
- Make AI agents work for everyone.
- Two layers. One company.
- Agent infrastructure and agent applications.
- Built in production. Open to developers.
- The same infrastructure. The same APIs.
Building AI agents?
Use the same infrastructure we run in production. Build with AIRuntime, AgentMQ, and TidyMessages.