#United States #New York #AI Agents #Financial Data #Daloopa

Daloopa's Groundbreaking Benchmark Report

Introduction

Daloopa has recently unveiled a pivotal report titled Benchmarking AI Agents on Financial Retrieval, which provides critical insights into AI performance in the financial sector. This comprehensive research analyzed 500 real-world financial questions and yielded striking results regarding the accuracy of various AI agent systems.

Key Findings

The report highlights that AI agents, fundamentally dependent on the data they access, displayed an astounding increase in accuracy rates when utilizing structured financial databases. Daloopa tested three prominent LLM-powered agent systems:
1. OpenAI's Agents SDK with GPT-5.2
2. Anthropic's Agent SDK with Claude Opus 4.5
3. Google's ADK with Gemini 3 Pro

When these agents relied on structured, auditable databases for financial data, their accuracy jumped to approximately 90%. This represents an impressive improvement of up to 71 percentage points compared to when they accessed data from public web sources, which are often deemed unreliable.

The Importance of Structured Data

One of the critical takeaways from this report is the recognition that higher accuracy levels are not solely the result of sophisticated models. Daloopa emphasizes that a significant component lies in the quality and structure of the data fed into these AI systems. High-quality, audit-ready financial data is therefore essential for effective financial retrieval tasks.

The study also elucidated that moving from 90% to 99%+ accuracy necessitates enhanced infrastructure surrounding the AI models. For example, issues like fiscal calendars and naming conventions often present hurdles. Daloopa observed that the AI agents tested performed better with US companies, most of which adhere to a December year-end, as opposed to non-US companies that frequently operate on different fiscal schedules.

Daloopa's Solution

Daloopa has positioned itself as a leader in addressing these data infrastructure challenges. Their service encompasses over 5,000 publicly traded companies globally and offers up to ten times more data points per entity than competing solutions. Furthermore, each data point is hyperlinked back to its source, enhancing transparency and auditability.

According to Thomas Li, CEO of Daloopa, “Our latest benchmark research underscores the necessity of equipping AI agents with high-quality data for FinRetrieval. Accuracy in AI-driven finance isn’t just a model problem, it’s a data access problem.” Daloopa’s offerings not only support the leading AI systems but also resolve prevalent issues faced by financial professionals in accessing reliable data.

Partnerships and Integration

In a bid to further enhance their technological capabilities, Daloopa has recently announced integrations with leading AI platforms. For instance, a new Model Context Protocol (MCP) connector with OpenAI's ChatGPT is expected to enrich workflows for ChatGPT users. Similarly, partnerships with Anthropic's Claude aim to extend the scope of financial services offered by Daloopa.

These integrations highlight Daloopa's commitment to providing essential data infrastructure that supports a range of analytical workflows. From assisting hedge funds in identifying fiscal trends to aiding researchers in generating comprehensive reports, Daloopa’s solutions are tailored to meet the evolving needs of the financial ecosystem.

Conclusion

In conclusion, Daloopa's benchmark report sheds light on crucial factors influencing AI performance in financial data retrieval. The necessity of structured, high-quality data cannot be overstated, particularly as reliance on AI in high-stakes sectors such as finance becomes more prevalent. By addressing these vital infrastructural needs, Daloopa is setting a new standard for accuracy and efficiency in financial AI applications. For more information or to explore a demo of their offerings, visit daloopa.com.