Compressed investment memo drafting time from weeks to just 30 minutes with Claude
Processed massive deal datasets containing thousands of documents using Gemini’s 1-million-token context window
Centralized all Claude and Gemini model inference securely within Vertex AI
Achieved 99.9% reliability for long-running workflows using a proven tech stack
Projected to scale inference spend 6x in 2026 due to rapid customer growth
Fifth Dimension uses Vertex AI to orchestrate a multi-model strategy with Gemini for large-scale document ingestion and Claude for high-precision reasoning. This approach allows them to securely and reliably compress complex commercial real estate workflows from weeks into minutes.
Fifth Dimension entered the market with a clear focus: use large language models (LLMs) to solve challenges in the document-heavy world of commercial real estate (CRE). Sitting at the intersection of finance and asset management, CRE professionals can be overwhelmed by data and documents, from retail leases to complex investment memos.
To address this, Fifth Dimension built a platform that offers many specialized AI capabilities for acquisition intelligence, asset optimization, and other common features such as reporting and communication. Its goal was to compress weeks of manual research and drafting for these groups into minutes while maintaining exceptional reliability.
Fifth Dimension needed a stable, scalable foundation to build its platform, as well as easy, real-time access to the most advanced models on the market. It ultimately chose to work with Google Cloud.
The commercial real estate market stands to realize a lot of benefits from generative AI given how document-intensive it is. With Google Cloud we can build tailored solutions that unlock more value for the industry.
Chen Wang
CTO, Fifth Dimension

For Fifth Dimension, innovation requires a stable foundation. The company intentionally chose to use a simple tech stack to support its modern AI offerings, prioritizing proven components for reliability and system uptime. It chose Google Cloud because it offered a more integrated solution to its small, agile team.
In addition to stability, security, and versatility, Vertex AI also ensures that input-output pairing remains consistent across versions as soon as a new Claude model release becomes available. This further accelerates our ability to test, onboard, and use Claude at scale.
Chen Wang
CTO, Fifth Dimension
To ensure 99.9% reliability, Fifth Dimension relies on the following Google Cloud infrastructure:

With this foundation, Fifth Dimension has stayed true to its commitment to use the best model for the job. The workflow integration within its platform brings together different models for complex, long-running tasks that sometimes last hours to intelligently match tasks to the model best-suited to a job.
For example, Fifth Dimension relies on Gemini Flash and Pro models for its one-million-token context window to process massive deal-related datasets that often contain hundreds or even thousands of documents.
When tasks require high-precision data extraction and reasoning, such as calculating rental income adjustments or generating code for Excel-based financial models, Fifth Dimension uses Claude Sonnet and Claude Opus models. Further, Claude Haiku is used as a specialized, cost-effective model for executing generated code on the fly.
Fifth Dimension was also impressed by Claude’s multimodal abilities. In testing, Sonnet 4.5 excelled at tricky tasks like reading complex diagrams and charts, and other details that had confounded other models.
Beyond that, Sonnet 4.5 shined with not just document understanding, but drafting capabilities as well, which helped further streamline Fifth Dimension’s review, analysis, and reporting process.
Fifth Dimension has also enjoyed some distinct advantages of running Gemini and Claude models through Vertex AI.
“Vertex AI supports long-running workflows that can sometimes last between 1 and 3 hours and have different models used and swapped out at various phases,” says Chen Wang, CTO of Fifth Dimension. “For example, we will sometimes use Gemini for initial large-scale document grouping, then switch to Claude for subsequent high-performance data extraction and reasoning. This is one of the many ways Vertex AI enables our “best model for the job” mentality.”
The impact of this multi-model approach has been transformative. Fifth Dimension has compressed the time to draft an investment memo from days or weeks to just 30 minutes. This efficiency has fueled growth, with the company closing major deals with top-10 U.S. asset managers and expanding its footprint with a 50/50 split between U.S. and international customers.
Internally, Fifth Dimension has improved AI adoption across its organization using Claude Code and Vertex AI. Every engineer at Fifth Dimension is an active Claude Code user. Engineers are generating more than 300,000 lines of code monthly and consistently accept 98% of suggestions.
Beyond engineering, account managers use Claude Code connected to BigQuery to run real-time data analysis using natural language.
“While some employees might have been apprehensive at first, Claude Code proved to be incredibly accessible for our non-IT teammates,” says Wang. “We’ve upskilled them, but it really didn’t take a lot of effort to help all of our employees leverage Claude Code to get the latest and greatest insights and data from sources like BigQuery.”
Vertex AI gives us the flexibility to orchestrate the best model for each task within a single, integrated platform. Our multi-model approach with Gemini and Claude has paid dividends for our clients.
Chen Wang
CTO, Fifth Dimension
Fifth Dimension anticipates continued exponential growth to push it towards 6x increase in inference spend in 2026, and has no intentions of slowing down.
Looking ahead, Fifth Dimension is focused on internationalization and helping customers build their own institutional "secret sauce" directly into the platform.

Fifth Dimension is a global AI technology company with offices in London, New York, and Singapore that specializes in automating complex, document-heavy workflows for the commercial real estate industry. The firm provides an AI native operating system for real estate to help global asset managers and owner-operators streamline investment and operational tasks.
Industry: Real Estate
Location: United Kingdom
Products: BigQuery, Cloud Run, Cloud SQL, Google Cloud Storage, Vertex AI, Claude, Model Garden on Vertex AI
About Google Cloud partner — Anthropic
Anthropic is an AI research and development company that creates reliable, interpretable, and steerable AI systems. Anthropic’s flagship product is Claude, a family of foundational AI models designed for enterprise applications.
