Capitalizing Alternatives Data on the Addepar Platform: Private Markets Benchmarking
Summary
- Private fund benchmarking: This video presents how Addepar leverages Databricks to build scalable benchmarks for private investments, turning vast alternative asset data into actionable insights.
- Data pipeline architecture: The Addepar Data Lakehouse, built on Databricks, demonstrates aggregation, filtering, and dynamic construction logic that processes millions of cash flows.
- Investor insights: The system enables clients to seamlessly compare their private fund investments against comprehensive benchmarks, supporting informed investment decisions.
Capitalizing Alternatives Data on the Addepar Platform: Private Markets Benchmarking

Talk by Ricky D'Sa, Lead Product Manager, Addepar
Addepar possesses an enormous private investment data set with 40% of the $7T assets on the platform allocated to alternatives. Leveraging the Addepar Data Lakehouse (ADL), built on Databricks, we have built a scalable data pipeline that assesses millions of private fund investment cash flows and translates it to a private fund benchmarks data offering. Investors on the Addepar platform can leverage this data seamlessly integrated against their portfolio investments and obtain actionable investment insights. At a high-level, this data offering consists of an extensive data aggregation, filtering, and construction logic that dynamically updates for clients through the Databricks job workflows. This derived dataset has gone through several iterations with investment strategists and academics that leveraged delta shared tables. Irrespective of the data source, the data pipeline coalesces all relevant cash flow activity against a unique identifier before constructing the benchmarks.
Here's more to explore:
Databricks named a leader in the 2024 Gartner® Magic Quadrant™for Cloud DBMS: https://www.databricks.com/resources/analyst-paper/databricks-named-leader-by-gartner
An open, unified approach to your data, BI and AI workloads: https://www.databricks.com/product/databricks-sql
See all the product announcements from Data + AI Summit: https://www.databricks.com/events/dataaisummit-2025-announcements
Chapters
0:00Introduction to Adapar1:29Alternative investment focus3:11Private fund workflows5:40Private fund benchmarks8:10Data engineering process8:55Data lakehouse architecture13:47Achieving data intelligence
FAQs
How large is Addepar's private investment dataset?
Addepar's platform manages $7 trillion in assets, with 40% allocated to alternatives, representing one of the largest aggregated private investment datasets available for benchmarking analysis.
What role does Unity Catalog play in private fund benchmarking?
Unity Catalog provides secure, governed access to sensitive investment data and enables delta shared tables that allow investment strategists and academics to collaborate while maintaining data security and lineage.
How does the data pipeline construct private fund benchmarks?
The pipeline aggregates and filters millions of cash flow transactions, coalesces them against unique identifiers, and dynamically constructs benchmark data through Databricks job workflows that continuously update for clients.
What business outcomes does this benchmarking deliver?
Investors can leverage private fund benchmark data integrated with their portfolio to obtain actionable investment insights, compare performance against peers, and make data-driven decisions on alternative allocations.
Full transcript
0:00 Introduction to Adapar
All right, thank you guys. Good afternoon. Thank you for joining me. I'm a product manager at Adapar. I'm excited to talk to you about private markets benchmarking. And thank you for joining me in one of the final sessions of the summit. I know you guys could have quit and gone back, but I'm glad you're here. And hopefully I'm going to insight you give you another insightful talk. I'm going to talk about benchmarking and how we leverage our partnership with data bricks to create a differentiated data and analytics offering.
We'll explore how Adapar provides deep insights and informed decision-making for our clients in the complex world of private markets. Before we begin, here's a quick legal disclaimer. All the information you see here is around sample andformational purposes and sample portfolio data. Here's what we'll cover today. We'll first start off with a brief introduction on Adapar, our story, our mission, and what drives us. Then we'll go over some of the capabilities around alternative investments.
Next, we'll focus in on the differentiated offering around private fund benchmarks and how we leverage our data lakehouse solution built on data bricks to power and create this data offering. And finally, we'll explore the underlying data platform that makes all this possible. Here's what we'll cover today. Adapar was founded in the wake of the global financial crisis over 15 years ago. We aim to help our investors answer simple questions around what do I own, where do I own it, and what's the impact on the markets on my portfolio.
1:29 Alternative investment focus
We strive to build an open architecture platform where we integrated with any sort of player in the ecosystem and we have grown exponentially over the past 15 years. We have hundreds of partners, over 450 different data providers that integrate inside of the system. And we have over 1,1200 clients, 1,200 clients, excuse me, 1,200 clients that represent 50 different global markets and over 100,000 users on the platform.
We serve the wealth management space across the board. Whether that's a family office, an RA, a registered investment advisor, an institutional firm, a private bank, our clients ultimately serve high net worth and ultra high net worth individuals that have complex portfolio structures with varying degrees of investment sophistication and importantly a high degree of confidentiality needed on their data. The assets on the platform are growing substantially fast with over 25 billion dollars added per week.
We have over 7 trillion of assets on the platform. This massive exclusive and fast growing data set requires advanced solutions and techniques. Hence our partnership with data bricks enables us to harness this data effectively in an aggregated anonymized fashion to create a differentiated data offering and deliver insights for our clients. Now let's review capabilities around alternative investments. Adapar was founded on the realization that we want to support our clients across all markets whether that's a public or private asset whether that's an Nvidia stock or whether that's a data bricks or adapar private security visualize that holistically within your portfolio all in one place given sophistication of our clients we see there's a whopping 40% of assets on the platform allocated to alternative investments If we look further into the segment of private funds, we have over 11,000 fund managers, 40,000 funds, and over 250,000 fund positions.
3:11 Private fund workflows
Further looking into the family office segment, you can see in green that's highlighted and further broken out private capital, hedge funds, private companies, other alternatives such as art or collectibles. Adapar truly holds a unique data set. This rich diverse data set affords us the opportunity to develop and create unparallel insights in the alternative investment space. I'll focus on the private fund investment workflow.
For those of you who are not familiar, a fund typically is an investment vehicle that comprises of a fund manager and investors that allocate money into that fund. They're considered LPS or limited partners. This is a pretty lengthy investment period and clients lock up their investment for a decade or plus. It's illquid and it's not as transparent to understand what that value of that asset is. Adapar assists the workflow from the pre-investment, the investment, investment period during that lengthy period, you're locked into that fund and even the post investment.
We have various solutions in place that assist our clients whether that's on the navigator side to assist with cash flow forecasting. Also, data management around ingesting the data across different types of documents in the private markets and analysis and reporting around performance measurement around your private funds. And finally, we'll focus more on private fund benchmarks, the newest offering. We understand our clients pain points and the space.
5:40 Private fund benchmarks
There's a lot of unstructured data. There's all diverse data sets pulled in manually. There's the challenge around commitment planning, forecasting what the allocation and commitments that needed to be allocated into that fund. And quite frankly, there isn't a standardization in this industry. There isn't a unique identifier or ticker or QIP associated to a private fund like we have in the public markets. Again, Adabar provides comprehensive solutions across these different challenges.
This brings us to our product spotlight on the product phone benchmarks to drive deeper performance insights. A core component of this offering is the quarterly performance benchmarks. As you can see, clients can leverage this information directly inside of their ADAR instance in context with their portfolio. Whether that's the specific fund, grouping of funds, or specific portfolio category, we break it down by the vintage, the strategy, or the substrate. and clients can see it across different time horizons, whether that's one, three, or five, or since inception.
We have data bricks workflow jobs in place that dynamically refresh the data at a particular cadence, and clients see the latest benchmark data inside of their instance. This isn't just about quarterly snapshots. It's about identifying trends and validating investment strategies. This granular benchmarked analysis empowers our clients to make more informed datadriven decisions. Beyond benchmarks, our diagnostic capabilities allow our clients to dig deeper and uncover some of the opportunities or areas of success.
They can see that against their specific fund investments or in context with their entire portfolio. We break it down by the investment strategy, the strategy allocation, excuse me, investment manager strategy allocation and understand the cortile rank against a cohort of different peers. This granular attribution helps our clients understand why performance is what it is, facilitating for targeted improvements and maximizing returns.
So how do we create all this? I talk about data bricks, but there's more to this entire workflow. The core platform ingests data from various sources systematically. We even pull in manual information and then ingest that into our lakehouse powered by data bricks. Normalize that information and then we run premputation against primitives. We have security mastering and validating the classification identification of these securities with thirdparty sources.
8:10 Data engineering process
And finally, we construct the benchmark through rigorous aggregation, anonymization, filtering, cleansing, and ensuring we have that right metric in place that clients can use against their respective portfolio. Now, let's delve deeper into the underlying architecture, the data platform. The Adap data lakehouse or ADL as we call it internally is the cornerstone of our platform. It is designed to ingest and unify a variety of different data types and data sources whether that's CRM data from Salesforce custodial data feeds that come in against a respective portfolio position and transaction configuration of a portfolio inside of their respective instance.
8:55 Data lakehouse architecture
Adapar allows you to flexibly configure that and associate against a respective portfolio data feed. And finally, the security master where we cross reference the information against trusted sources and unify the client to respective data feed ultimately the ownership structure and the portfolio within their instance. A key element of ADL is premputation. The why is critical. Our objective is to deliver fast queries at low cost ultimately unlocking new product domains on that increasing amount of data set.
As I mentioned earlier, our data is growing fast and so this is a core element as part of our priv benchmarks. As you can see in this chart, on the-fly computation where calculations happen on the fly every time the query is made, the response time is linear as the volume of data increases. This can lead to frustrated users, slow dashboards, and ultimately lost opportunity. Premp computation, however, flattens that curve and significantly reduces the burden.
There's consistent rapid performance as the data scales. So, what do we premputee? Our philosophy is as little as possible but as much as necessary and what we call the primitives. These are the fundamental building blocks the core data elements within a portfolio structure. The design trade-offs are where we have more data storage and network costs. But we believe this is far less than the burden required in computationally running these metrics from scratch.
Additionally, Adapar's inherent flexibility means you can have a myriad of different ways and ownership structures that are directly, indirectly or partially owned. So that computation can be very burdensome. Some precomputed examples include the value, the cash flow, the ownership structure, and the performance metrics around standard metrics used in private funds like IRRa, TVPI, DPI. By having these core components calculated, pre-calculated, our system can assemble and present complex insights almost instantaneously, no matter how detailed that query is.
Following the concept of preomputation, let's examine an example. The table on the left shows you a simple portfolio structure, an ID name match where Tony Stark has a brokerage account and the unaligned security is Stark Industries and table on the right is the closure table. But this isn't just a direct parentto child relationship. It is a precomputed map of all the possible paths and depths within an ownership structure.
As you can see, Tony Stark directly against their security is a depth of two. The closure table is crucial for navigating the often deeply nested and complex hierarchies found in private markets and the inherent flexibility that Adapart provides you. This can be very detailed. By having these primitives pre-calculated, we're able to instantaneously produce the results and calculating the IR TVPI across different time frames against the respective portfolio.
Now imagine trying to aggregate this across dozens of funds all in real time. The closure table allows ATAPAR to perform these aggregation and calculations with incredible speed and accuracy providing for a unified multi-resolution view using a single source of truth against our security master. It is a foundational piece that enables deep analysis without sacrificing for performance. This slide encapsulates it all. What we're trying to achieve is datadriven investment decision-m attaining financial intelligence starting from the bottom where we have the data set aggregate across different sources private and public markets. then constructing that into data in context where clients can identify trends, outliers and now when they use that against their respective portfolio, they have the knowledge in place to then validate an assumption or or hypothesis made on their portfolio.
13:47 Achieving data intelligence
At the Apex, the intelligence is that decision that can be made with this knowledge. Now whether that's rebalancing your portfolio, modifying your pacing model or modifying the entire investment strategy of a portfolio. The product from benchmarks is built on this framework. Next, we plan on creating a self-service capability and platform and hopefully using some of the new tools and solutions that data bricks announced yesterday and today where we can enable a faster deployment and creation of these data products and analytics and investor tools for our clients.
Data bricks enabled the first data product and we're excited for building more and maximizing the impact of the world's capital. Thank you.
Learn more about the Databricks Data and AI platform.
The information provided herein is for general informational purposes only and may not reflect the most current product capabilities or configurations.