← Sticker Wall AGENT · DETAIL

Dust - Helping enterprises build knowledge base AI assistants using a French RAG platform, secured $40M in funding

Workflow: Independent service providers connect enterprise data sources (such as Notion and Google Drive) in the Dust console, cre

AGENT

Key Fields

FIELD STAMPS

IndustryContent / Creator Economy

RegionEurope（欧洲（法国））

ScaleSME

ChannelOnline

🔧 Workflow

Independent service providers connect enterprise data sources (such as Notion and Google Drive) in the Dust console, create a queryable internal knowledge assistant via RAG, and configure permissions and token usage. After employees ask questions daily, the system automatically retrieves relevant materials and generates answers. Service providers regularly review usage logs and optimize prompts and data indexes. Additionally, Dust supports integration with collaboration tools like Slack, allowing employees to invoke agents directly within chat windows, while service providers use the backend to analyze high-frequency questions and iteratively update the knowledge base content. For enterprises requiring cross-departmental workflows, service providers can configure multiple agents to collaborate—for instance, one responsible for retrieval and another for summarization—ultimately delivering unified outputs to designated channels.

🛠 Setup Requirements

Basic API and prompt engineering skills are required, along with the ability to use the Dust console to connect data sources. Initially building a test agent takes about 1 to 2 weeks, after which it can be rapidly deployed across different enterprises. For non-technical backgrounds, Dust offers a no-code canvas and ready-made templates, enabling basic agent setup through drag-and-drop. It is recommended to start with a single vertical industry, build a standardized industry knowledge base template, and then horizontally replicate it for similar clients to significantly shorten the delivery cycle per order.

🧰 Toolchain

🔧 Dust
🔧 OpenAI API
🔧 Notion
🔧 Slack
🔧 Google Drive

💰 Revenue

Individual revenue has not been disclosed. The platform charges based on seats and tokens. Dust itself secured $40 million in funding and has 300,000 agents live, indicating strong enterprise willingness to pay. For enterprise clients, Dust bills by seats and token consumption, allowing independent service providers to charge additional implementation and maintenance fees on top of this, establishing recurring revenue. Based on typical pricing in the European market, monthly maintenance fees for a single enterprise can range from hundreds to thousands of euros, depending on the service provider's own capabilities.

💸 Cost

Using Dust requires paying platform subscription fees and model API token fees, with specific amounts subject to official website pricing; beginners can start using free tiers for validation. It is recommended for independent service providers to list model API costs separately in their quotes to prevent cost overruns caused by high-frequency client queries. Starting with small-scale pilot clients can effectively control token consumption before scaling up after the workflow is proven.

⏱ Time Investment

About 20 hours per week, with initial setup taking around 2 to 3 weeks. Afterwards, weekly monitoring of usage and prompt optimization is sufficient, with reduced effort required during the maintenance phase.

🚀 Getting Started

First, register an account on Dust, create a test agent, connect public knowledge documents, and experience the RAG retrieval performance. Then, target local small and medium-sized enterprises to pitch internal knowledge base AI assistant services, charging monthly fees based on seats and token usage. You can also join the Dust community or official Discord to get the latest case studies and templates, accelerating the learning curve. After securing your first benchmark case, leverage industry communities and content marketing to acquire more clients.

🔑 Keys to Success

✅ Data ingestion quality directly determines answer accuracy
✅ Design pricing packages based on token usage to control costs
✅ Focus on niche industry knowledge bases to establish benchmark case studies
✅ Continuously iterate knowledge base content to maintain answer freshness

⚠️ 风险

⚠️ Enterprise data security and compliance requirements are high, requiring careful handling of permissions and privacy
⚠️ Changes in platform pricing or model API costs may squeeze independent service providers' profit margins
⚠️ Large tech companies and low-code platforms also offer similar features, requiring deep industry expertise and fast service response to win

📌 Real Cases

📌 Dust secured a $40 million Series B funding round, with 300,000 agents live on its platform, used by multiple European enterprises for internal knowledge retrieval, as reported by FrenchWeb highlighting its accelerated multi-agent system deployment
📌 Uclic reported that Dust raised a $40 million Series B funding round led by Sequoia, emerging as a representative French enterprise-grade AI agent platform
📌 Cocoloop reported that the number of agents on the Dust platform reached 300,000, demonstrating sustained growth in enterprise client demand for internal knowledge assistants

Sources