Transforming research through AI-enabled collaboration
In today’s capital markets, timing is everything. Clients expect differentiated insight at speed, and analysts are under increasing pressure to deliver expert commentary across broader coverage in near real-time.
At RBC Capital Markets, we’ve always been proud of the quality of our research. But in a rapidly evolving environment, quality alone isn’t enough – we needed to reimagine how we could deliver insights with the same depth and rigor, while significantly accelerating time to market.
That belief sparked one of our most transformative innovations, the creation of Aiden QuickTakes: a proprietary AI application built to accelerate and scale research coverage while preserving the human expertise our clients value most.
Rethinking the analyst workflow
Traditionally, producing a post-earnings research note could take an analyst upwards of 45 minutes. The workflow was highly manual – sourcing the announcement, extracting key figures, writing and formatting the note, then publishing it. Multiply that across dozens of companies, and it created a bottleneck.
Analysts wanted to spend more time on analysis and delivering client insights than on process. And in a world where even a 30-minute delay can mean missed market opportunities, that model wasn’t sustainable.
We asked ourselves: what if we could automate the repetitive and amplify the insightful? What if we could streamline the creation of earnings summaries while preserving the expert voice and judgement of our analysts?
That’s when QuickTakes was born.
Creating a secure foundation for AI innovation
Delivering on this vision required more than just building an AI model. It required a secure, scalable data foundation – one that could unify massive volumes of structured and unstructured data, support real-time processing, and provide end-to-end governance.
This is a highly challenging undertaking both in complexity and accuracy (requiring 99%+). The technology and innovation required goes beyond traditional AI and Intelligent Document Processing (IDP) strategies.
For that, we worked with Databricks which provided a unified, cloud-native platform that allowed us to move seamlessly from data ingestion to analysis to content generation – all within the same environment.
Two of the most critical components of this setup are Databricks’ Unity Catalog—an open governance layer that provides centralized security and oversight for data and AI—and Mosaic AI, which delivers end-to-end AI capabilities within the Data Intelligence Platform. It allows us to consistently manage access, security and compliance across teams and systems, even as we scale. This governance capability is especially important in a regulatory environment where trust and transparency are non-negotiable.
Databricks’ native support for model development, orchestration, and deployment meant we could embed generative AI directly into the research workflow. Combined with high-performance infrastructure for data processing and inference, we now have the agility to evolve quickly, without compromising quality or control.
Also, this platform doesn’t just enable our current capabilities, it evolves with us. The continued investment in open standards, scalable infrastructure, and enterprise-grade AI gives us the confidence to expand without limitation.
“A modern Data Intelligence Platform is essential for organizations seeking to unlock the full value of their data and drive innovation with AI. Our collaboration with RBC Capital Markets demonstrates how industry leaders can harness AI to significantly boost employee productivity, without compromising the quality of the insights or the rigorous standards of security and governance required in highly regulated environments.”
Naveen Rao, VP of AI, Databricks
QuickTakes in action
QuickTakes was designed to streamline what analysts spend the most time on, without compromising what clients value most. When a company announcement goes live, the system automatically ingests the release, extracts financials, and produces a first-draft research note. Within minutes, the analyst has a high-quality draft in hand, ready to refine, add insight, and publish. Human-in-the-loop remains a crucial part of our analysts process.
“QuickTakes doesn’t replace human expertise – it enhances it. It frees analysts to focus on what clients’ value most: insight.”
Bobby Grubert, Global Head of AI and Digital Innovation, RBC Capital Markets
The result is a faster, more agile process – one that has reduced turnaround time by over 60%, expanded analyst coverage, and reclaimed thousands of hours for higher-value activities. What once took 45 minutes now takes just seconds and with “human-in-the-loop” takes 15 minutes from announcement to publishable draft.
Most importantly, our analysts remain in full control. QuickTakes doesn’t replace human expertise, it enhances it. It frees analysts to focus on deeper commentary, sharper takes, and richer client conversations.
“This isn’t just about speed. It’s about redefining how insight is created, scaled, and delivered to our capital markets clients.”
Michael Tran, Global Head of Digital Product Innovation, RBC Capital Markets
Expanding the impact
The Aiden platform – our broader AI ecosystem – now powers multiple aspects of our business, from accelerating document preparation in investment banking to summarization, content generation, new insights and data creation for our teams. These aren’t isolated wins; they’re interconnected. They’re part of a deliberate shift toward an operating model where AI, data, and human judgment work in tandem.
Together, we have built a framework that supports not just today’s needs, but tomorrow’s ambitions. That includes exploring new AI models, embedding smarter analytics into client interactions, and continuing to reduce friction across the enterprise.
A model for innovation
What makes QuickTakes truly special isn’t just what it does – it’s how it was built. It represents the best of what happens when cross-functional collaboration meets strategic technology collaboration.
Our analysts, engineers, product leads, and data scientists worked hand in hand with Databricks to develop, test, and refine every element of the system. It was a collaboration in the truest sense, shaped by user needs, refined through real-world application, and designed to scale.
