holaboss
holaboss builds agent-native infrastructure for service companies: a system that lets AI agents handle the handoffs from winning clients through delivering work. Its runtime, model gateway and research stack are intended to let companies operate their business on agents while retaining control of data, tools and costs.
At a Glance
- Solo operators and small agencies
- Growing agencies
- Service companies and professional-services firms
- Enterprise organizations
- +2 more
AI Tools by holaboss
(1)holaOS
Local Agent Workspace Desktop App
Discussions
No discussions yet
Be the first to start a discussion about holaboss
Latest News
holaOS repository README update documents the local-first agentic workspace, shared memory, integrations, multiple agents and model choice.
holaOS release event recorded on the repository; Holaboss search results identify the product launch as holaOS Beta 0.1.
Initial commit for the public holaOS open-source agentic workspace repository.
Holaboss website presents the expanded agent-native infrastructure product for service companies, including six operating chapters and the holaOS/holaProxy/Neo Lab stack.
Products & Services
Agent-native infrastructure for service companies, covering business website building, AI employees, customer management, delivery management, a Company Brain and reusable service templates.
Open-source, local-first agent-native desktop/workspace runtime. It supports macOS, Windows and Linux, shared memory, apps, chat tools, browser and local files, MCP, integrations, multiple models and agents such as Claude Code and Codex, with BYOK and self-hosting.
A model and integration gateway providing one API across models, unified credits, token metering, workspace-level cost tracking and support for bring-your-own keys.
A community feed and recipe platform for sharing and one-click installing agent skills, apps, MCP servers, integrations and workspace configurations.
Market Position
Holaboss positions itself as agent-native operating infrastructure for service companies rather than another standalone chatbot or conventional CRM. Its differentiation is the end-to-end operating model across website, sales, customer record, delivery, institutional memory and service templates, combined with a local-first open-source runtime, broad integrations/MCP support, model choice and usage-cost metering. It competes conceptually with CRM and professional-services operations suites, agency automation tools and closed AI-agent/workspace platforms.
Leadership
Founders
Haoxiang (Harry) Wang
Founder of Holaboss; LinkedIn search results identify him as a Palo Alto-area founder with a University College London economics and business background and venture-capital experience.
Executive Team
Haoxiang (Harry) Wang
Founder
Founder of Holaboss; LinkedIn search results associate him with University College London and venture capital.
Founding Story
Holaboss was started around the idea that AI agents should operate as a company's working layer rather than as isolated chat tools. The initial vision was an AI-native operating system for individuals and organizations, evolving into infrastructure that closes the handoffs between lead capture, CRM, sales, delivery and repeatable service execution.
Business Model
Revenue Model
Subscription plans for hosted holaboss workspaces, plus usage-based credits for brokered model and integration calls; hosted plans also charge a percentage of customer collections on Starter and Growth. Enterprise pricing covers custom, white-label, dedicated-cloud and on-premise deployments. holaOS can be self-hosted, while hosted workspaces and commercial distribution of the modified- Apache-licensed runtime are monetized separately.
Pricing Tiers
1 seat; approximately 500 included credits; 2% commission on collections; 1 published client app; subdomain hosting; private Company Brain.
Up to 5 seats; approximately 6,000 credits from day 8; 1% commission; up to 3 published apps; custom domain; Slack and email channels; white-glove setup; 7-day trial.
Custom seats; approximately 25,000 included credits; no commission; unlimited apps and domains; full white-label/eject; every channel including WeChat; white-glove setup and SLA.
Credits do not expire; model calls through Holaboss are approximately 1 credit per 2,000 tokens, while integration calls are metered per call.
Target Markets
- Solo operators and small agencies
- Growing agencies
- Service companies and professional-services firms
- Enterprise organizations
- Organizations needing customer-facing or employee-facing embedded agents
- Developers and enterprises building self-hosted agent workspaces
- Running service businesses on AI agents
- Lead capture, qualification and sales follow-up
- Client/customer relationship management without manual CRM updating
- Project and engagement delivery management
- Turning repeatable agency or professional-service work into service templates
- Building and operating local-first agent workspaces