Liquid Web is a leader in managed web hosting services, providing reliable hosting solutions to businesses worldwide. The company is committed to leveraging AI to enhance customer experiences and streamline internal operations.
With everything happening in AI, Liquid Web wanted to stay ahead of the curve in the web hosting ecosystem. By making the best use of generative AI, they wanted to boost their developer productivity, reduce their grunt work and make their support infrastructure robust and reliable.
AI Challenges Before Langbase
Before choosing Langbase, Liquid Web faced key challenges in implementing AI solutions:
- Lack of transparency and control: Existing AI tools offered limited visibility into backend processes, making it difficult to troubleshoot issues or optimize performance.
- Hallucinations and unreliable responses: Previous AI implementations frequently generated inaccurate or misleading information, eroding trust. Liquid Web engineers kept fighting the LLM pre-training based responses which led to a lot of hallucinations in support answers.
- Steep learning curve: Some AI tools required extensive domain expertise and customization, creating a high barrier to adoption.
- Hard to evaluate and refine AI outputs: Liquid Web needed a way to validate AI-generated responses before deployment to ensure accuracy and consistency. Developers experience an after-thought: Hard to version, backup, and restore agents.
Langbase helped Liquid Web build fast, composable, and reliable AI agents
Langbase stood out as the ideal AI solution for Liquid Web due to its composable AI primitives, which can be easily deployed serverless. Within a week Liquid Web team was shipping production quality agents that were solving real customer support problems.
- Reliability and accuracy: Langbase memory agents offered superior accuracy for AI-generated customer support responses. It’s easy to understand how memory agents make a decision, developers were able to see exactly what chunks were used to answer questions, getting insights into how memory and pipe agents reasoned over a predefined knowledge base, minimizing hallucinations to less than 5%.
- Scalability and ease of deployment: Langbase helped Liquid Web deploy and scale quickly. Building AI agents, forking, versioning, restoring, experimenting, and working with all the stakeholders in the AI studio — provided an excellent developer experience.
- Composable and Transparent Liquid Web enjoys how Langbase puts them in control by enabling extensive customizability to build AI agents and plug them in any new or legacy system without buying into a framework like abstraction.
- Transparency and with zero-vendor lock-in: The built-in memory testing feature allows developers to see exactly which knowledge chunks are being retrieved, offering unprecedented visibility. To top it off, being able to switch between 250+ LLMs without any API changes allows Liquid Web to easily switch to the latest frontier model for the task at hand with the right speed, cost effectiveness, and powerful intelligence.
The real breakthrough here is how easily we can test RAG via Langbase memory agents—actually seeing which chunks get retrieved for specific queries. That kind of visibility just isn’t available with other providers. Typically, you’d need specialized vendor knowledge, framework-centric coding, or custom actions to get the same result. But with Langbase, there’s next to no overhead or prior expertise needed. The straightforward setup, developer-friendly testing tools, and built-in version control instantly give it a huge plus for anyone building AI-driven apps. Langbase basically turned me into an AI engineer overnight. The learning curve is practically zero thanks to the clear explanations and clean UI. And being able to talk directly with real engineers on their support team makes the entire experience even smoother.
— Katie Jackson, AI Development & Infrastructure, Liquid Web
How Liquid Web is leveraging Langbase
Liquid Web is using Langbase to tackle:
Internal AI Support Agent
Liquid Web leveraged Langbase to power an internal AI agent built on top of Slack that retrieves information from proprietary documentation. Given the sensitive nature of this data, Liquid Web needed an AI solution that could operate securely with deep customization, visibility, and transparency.
Competitive Intelligence & RAG Testing
By using Langbase’s rapid indexing capabilities, Liquid Web was able to ingest competitor documentation and test AI-generated responses against it. This allowed for fast prototyping and validation of AI-driven tools.
We went to a competitor’s site, downloaded their publicly available documentation, and tested it with Langbase. The AI agent remained consistently accurate, never veering from the provided data. This gave us enough confidence to move fast on day one.
— Stephen Gregorowicz, Internal Tooling lead, Liquid Web
Scalable AI Infrastructure Langbase's composability and modular AI primitives have empowered Liquid Web’s engineers to design a flexible, AI-powered stack that can be deployed across multiple systems with ease. The ability to backup and restore AI models quickly ensures business continuity.
Results with Langbase
Since integrating Langbase into its AI workflows, Liquid Web has experienced significant improvements:
- 30% increase in efficiency across AI-driven processes.
- One workday per week saved per user due to improved response accuracy and automation.
- 95% successful customer queries resolved without any hallucination exceeding initial estimates.
- Minimal hallucination rates, ensuring reliable and consistent AI outputs.
- Faster deployment cycles, allowing rapid testing and iteration of AI-driven support tools.
Continuous Refactoring and Moving Ahead
Liquid Web’s adoption of Langbase is proving to be a game-changer for their team. With the confidence gained from developer productivity and increased system efficiency, Liquid Web is ready to initiate new projects with Langbase to reduce their tech debt and decompose their legacy architecture.
Stephen and Katie are committed to continue the following with Langbase:
- LLM advancements tailored for enterprise use cases.
- Unbundling Langbase’s memory agent to gain even more granular control over AI-driven workflows.
- Expanding AI-driven automation to product and engineering teams.
We’ve only scratched the surface of what’s possible with Langbase. The partnership is just getting started, and we’re excited for what’s next with AI primitives and multi-agent capabilities.
— Stephen Gregorowicz, Internal Tooling lead, Liquid Web
With 95% AI response success rates, one workday saved per user per week, and minimal hallucination rates, Liquid Web has set a new standard for enterprise AI adoption.
Langbase continues to be a trusted AI partner, enabling Liquid Web to push the boundaries of AI-driven innovation while maintaining full visibility and control over its systems.
Learn more about Langbase and how you can improve your engineering operations by signing up for a demo today!
