Get clear insights about your portfolio
Ask plain language questions about fund and portfolio data
Investors and the IR team get answers when they need it, and fund operations gets time back to focus on in depth analysis.
Fully customized for your funds and your data
Spine pulls from your databases, and your model is trained on standard investment acronyms and terminology, as well as any inside metrics and terms specific to your business.
Be confident in self-serve reporting
Provide custom instructions for complex questions
How we deliver an AI analyst that understands your business
1
Onboard
Identify the scope of your AI analyst, gather domain context to create custom embeddings and representative examples to train against.
2
Test and refine
Test the initial version with your live data, evaluate for consistency, and refine based on human feedback.
3
Deploy and monitor
Deploy to your users, and monitor on an ongoing basis to ensure accuracy.
Constant pings to fund operations to get LPs details about fund performance, or give investors updates on portfolio companies
Dozens of hours per month spent on analysis for ad-hoc questions for IR teams and investors
Backlog of questions to work through following busy periods, like quarterly reporting, fundraising or due diligence
Off the shelf LLM solutions are a black box, that can’t be trusted, because there’s no visibility into how they produced their answers
Inaccurate Al chats, prone to hallucinations and don’t understand key terminology, acronyms and concepts that are critical to private market investing
Fund operations gets back their time and focus for the work that needs detailed analysis and expertise, like valuations, forecasting, and portfolio construction and optimization
From 4+ hours of human analysis for complex ad hoc questions, to less than 10 minutes. And make the same process repeatable in the future.
Full transparency into where the data came from, so you can be confident enough to make investment decisions or share answers with LPs
The ability to monitor and refine the responses investors and the IR team receive as needed
Flexible, reliable, and secure by default
Works with any data store
We can accommodate any structured data. Connect your CSVs, APIs, and databases to our product.
Explainable and customized
Our proprietary DPU allows for explainable, individually tested and domain fine-tuned modules that work at scale
Truly natural language
Our semantic layer allows us to accurately interpret and handle natural language questions across multiple large datasets
Rigorous evaluation
Combined automated and human evaluations give us an objective way of measuring performance — both before and after we deploy the analyst.
Secure by default
Isolate data between users with our built-in tenant isolation. Access is scoped to the user's permissions, just like the rest of your front end.
Founded by a world-class team
Our team has helped bring to life over 40 AI products. We have 100+ citations, 500+ Github stars, and 2 patents.