How predori cut operational costs over 80% with an AI-Powered Patent Intelligence Platform

Summary

For a growing AI-Tech company, every dollar and every engineering hour counts. predori, an ambitious patent intelligence software, found itself burning cash on a costly, inefficient system that wasn't built for modern AI. By switching to Weaviate's managed vector database, they reduced their operational costs by over 80%, freed their team to innovate, and cleared the path for building advanced agentic RAG systems that will define their future.

The Challenge

Migrating from a Costly and Inefficient System

predori was founded in 2020 with a clear mission: make patent research simple and accessible. From the start, they knew semantic context and vector retrieval were the future. However, in the early days before vector databases became mainstream (late 2022), the team was forced to build its capabilities on a Lucene-based AWS service.

This setup was not designed for the mathematical problem of approximate nearest neighbor search. To make it work, the team had to keep everything in RAM and over-provision hardware, leading to a system that was "super expensive and just not good." For CTO Konstantin Schmitz, the choice was clear: find a scalable, purpose-built vector solution or risk being buried by infrastructure costs.

"We were trying to squeeze a vector database into what is not really a vector database, and it was super expensive and just not good. It was just the wrong technology, a complete overkill for the mathematical problem of approximate nearest neighbor search."

- Konstantin Schmitz, Co-founder of predori

The Solution

A Production-Ready Patent Intelligence Platform Built on Weaviate

The team knew they needed a purpose-built vector database. After experimenting with a range of self-hosting options like Facebook's Faiss and evaluating managed providers like Milvus and Pinecone, they chose Weaviate Cloud in 2021. The decision came down to three key factors that were critical:

• Predictable Pricing: Weaviate's transparent pricing and sandbox environment gave the team the confidence to experiment and scale without worrying about runaway costs.
• Operational Freedom: As a fully managed solution, Weaviate Cloud handled the operational heavy lifting of hardware management. This allowed predori's team to stop being sysadmins and get back to what they do best: building their software.
• Ease of Migration: The team was able to seamlessly switch from their previous AWS infrastructure to Weaviate Cloud. Konstantin described the process as a "natural" transition that felt like a "drop-in replacement," simply swapping out the old client connection for the Weaviate Python client.

Results

Cost Reduction, Faster Innovation

• 80%+ Cost Reduction: By moving away from the Lucene-based AWS setup, predori immediately cut their operational costs by a factor of 10.
• Redirected Engineering Focus: With a dedicated vector database partner, the engineering team was able to “iterate on the product side and develop new features in the RAG space”, without the distraction of building their own solution, a process that would have "greatly delayed our time to market." - Konstantin Schmitz, Co-founder of predori.
• Expansion into Agentic RAG: Weaviate’s scalable foundation has supported predori to move beyond basic search to develop advanced agentic RAG systems. They could now focus on building new, high-value features instead of constantly battling their infrastructure.

What's Next?

The partnership is now focused on optimization. predori is currently implementing Weaviate’s multi-tenancy and vector compression features to further optimize their service in light of its growing user base and the continued expansion of its capabilities. For them, the value isn't just in the technology, but in the partnership. As Konstantin put it, the experience has been "very pleasant."