Product descriptions:
Franklin Templeton manages over $1.6 trillion* in assets across traditional, alternative, and digital investment products, serving Financial Professionals in more than 150 countries. As the products available to Financial Professionals multiplied, so did the demands on the firm’s analysts and distribution teams to deliver clear, insight-driven guidance. By building portfolio analysis and commentary solutions with Agent Bricks on Databricks, Franklin Templeton is enabling faster response to investment-related inquiries, greater coverage, and tailored collaboration — serving more Financial Professionals with greater clarity and confidence.
*as of 9/30/2025
Increasing complexity, manual bottlenecks, and the need to scale
The expanding universe of mutual funds, ETFs, digital assets, and alternative investments made it challenging to deliver timely, customized insights. “My focus is helping our Financial Professionals navigate an increasingly complex world of investment choice,” said Mark Nigro, CFA, CAIA, Head of Portfolio Selection and Research Services. Legacy processes limited capacity for analysts, a team of just seven who produced manually authored, highly researched notes for Franklin Templeton’s distribution teams and their financial advisor clients. “It’s resource-constrained everywhere,” added Colin Zimmerman, CFA, Lead Data Scientist. “We were able to support 200 products with analyst-generated commentary, but we always left opportunities on the table.”
Standard dashboards and shared libraries couldn’t keep pace, and early experiments with foundation models also fell short — they generated text but lacked grounding in the firm’s data. For a regulated business, these models were unusable in production. The team needed a governed, explainable AI framework to deliver analytical support faster across more products by combining proprietary data and analyst expertise. As Mark explained, “We wanted to move from dashboards and stale notes that left our distribution teams disengaged to a solution where they can get actionable, custom analysis without waiting days for an email response.”
Purpose-built agents with secure, accurate, real-time insight
To address these challenges, Mark Nigro and Colin Zimmerman partnered to combine investment insight, portfolio construction, client preferences, and data science. The result is SIGNALS, an internal application powered by Databricks that delivers these capabilities to analysts and distribution teams.
The firm achieved a significant milestone by using Databricks Agent Bricks. Mark and Colin unified both their proprietary fund scoring models and critical unstructured documents, such as prospectuses, fact sheets, and portfolio manager commentary, within Unity Catalog and the Databricks lakehouse architecture. “With Agent Bricks, we can blend our unique algorithms with up-to-date enterprise documents and automatically generate insightful product notes that sound like our experienced analysts,” shared Colin.
“What’s exciting about this is that it started as a business challenge,” said Mark. “Colin and I worked together to combine investment insight and data science in a way that gives our analysts and sales organization a real tool they can trust and use every day.”
Rather than deploying a generic chatbot, Colin purpose-built several agents with Agent Bricks, including a custom LLM agent trained on 200 analyst-authored notes, and two extraction agents that process PDF fund fact sheets, commentary, and other materials into structured, machine-readable formats. This approach ensures analyses are always based on the latest internal and market data, never generic or outdated sources.
All data powering these agents is centrally managed in Unity Catalog, minimizing business risk and reducing hallucinations by grounding every response in approved enterprise logic and inputs. “If we didn’t have those two things — the scoring algorithms and the unstructured enterprise documents — the agent wouldn’t have the level of detail and granularity it needs to replicate a human expert,” Colin explained. This foundation transformed their AI from interesting prototypes into deployable, auditable insights that analysts could rely on. Agent Bricks integrated evaluation tools further enable the team to iterate rapidly, tuning prompts and instructions so the AI commentary consistently meets length, clarity, and compliance requirements. “The evaluation loop made it simple to tune our guidelines and ensure outputs looked like an experienced analyst’s work — not just surface-level stats,” noted Colin. The combination of multi-agent orchestration, robust governance, and ongoing evaluation means Franklin Templeton can deploy new capabilities quickly, with the confidence that every insight is accurate, reliable, and audit-ready.
These tools augment rather than replace analyst expertise. By automating repetitive analytical steps, they free analysts to interpret results and engage Financial Professionals on more complex business opportunities, extending human insight through trusted AI collaboration.
Broader coverage, faster decisions, and measurable savings
With Agent Bricks, the firm can deliver custom analysis for far more products in far less time, expanding coverage from 200 to hundreds of funds and ETFs and helping distribution teams surface the best fit for each Financial Professional, faster.
The new capability is already live through SIGNALS, an internal application used by analysts and field teams. “SIGNALS helps our teams work faster and smarter,” said Mark. “It gives them access to consistent, data-driven portfolio insights and product comparisons they can immediately use in conversations with Financial Professionals.”
Mark released SIGNALS to more than 300 salespeople in July. The shift has been transformative. Sales teams can now instantly verify opportunities. Analysts save 2+ hours per week, reclaiming more than 15 collective hours each week. During the first months of the expanded rollout, field teams reported uncovering $15 million in new product opportunities. Usage is increasing, ensuring that Financial Professionals receive timely and relevant insights. “We now have access to instant replacement options and additional portfolio considerations that previously took days, insights that hold up to analyst-level scrutiny,” said Mark.
The solution is proven and extensible. “We’re not just solving existing bottlenecks — we’re building a blueprint to serve broader firm partnerships, whether for new geographies, channels, or product types,” Mark observed. Mark expects AI elements of SIGNALS will be leveraged for their Portfolio Analytics Tool, the firm’s web-based portfolio and market research platform accessible through FranklinTempleton.com.
Franklin Templeton’s Agent Bricks journey demonstrates how domain-specific, governed AI can accelerate business operations, driving better information in less time while adhering to compliance and business standards. “Databricks has helped us leverage internal and market data to greatly scale our ability to support Financial Professionals as they make investment decisions,” Colin concluded.
Please note, Franklin Templeton maintains an ownership interest in the artificial intelligence platform referenced in this material. As a result, Franklin Templeton may benefit from the use or promotion of this technology.