From Days to Minutes: How High Meadow Solutions Put an Institutional Asset Manager’s Data to Work with Claude

A High Meadow Solutions customer story

Summary: A multi-billion-dollar institutional asset manager wanted its investment and operations teams to get trusted answers from their own data without waiting on analysts or risking the hallucinated numbers that make generic AI tools unusable in finance. High Meadow Solutions (HMS), a Services Partner in Anthropic’s Claude Partner Network, co-delivered, in partnership with Gluon Digital, the solution: AskBarry, HMS’s own AI data analyst built on Anthropic’s Claude, running on a purpose-built, auditable data foundation entirely inside the firm’s own cloud. Today, anyone at the firm can ask a question in plain English and get an answer, with the underlying query, source, and lineage attached, in minutes.

The Challenge

Institutional asset managers run on data, but rarely with it. The numbers that matter, positions, flows, client and consultant relationships, pipeline, performance, live across a patchwork of systems: a Salesforce CRM, portfolio and accounting platforms, and asset-management-specific tools that don't talk to one another cleanly.

For this firm, that meant every non-routine question turned into a project. An analyst or relationship manager who needed a cut of the data had to file a request and wait for a small, overloaded data team to write a query or build yet another dashboard. Answers arrived in hours or days, often after the moment that prompted the question had passed.

The obvious fix, a general-purpose AI chatbot, was a non-starter. In an industry where a single wrong figure can misinform an investment or compliance decision, a model that confidently invents numbers is worse than no tool at all. Leadership also had hard requirements that off-the-shelf AI couldn't meet: sensitive data could never leave the firm's environment or be used to train an outside model, and access had to respect who was allowed to see what.

The firm needed self-service answers that were **fast, trustworthy, traceable, and secure**, not a black box.

The Approach

High Meadow Solutions and Gluon Digital co-delivered that foundation and deployed AskBarry, HMS’s productized AI data analyst, on top of it. The work followed HMS's standard five-phase build: consolidate the firm's CRM and source data into a governed cloud data warehouse, model an AI-ready dataset with strict metric definitions and data-quality controls, build the connective layer and skills that let Claude reason over that data, add the security and access controls, and roll AskBarry out to users at their desks.

The design rested on three principles that directly answered the firm's concerns:

A purpose-built datafoundation. Rather thanpoint a model at raw, messy systems, HMS architected a clean, unified data layer with the firm's business definitions encoded into it. Because the metrics are defined once and governed centrally, AskBarry never guesses what "AUM" or "net flows" mean. It asks the same governed source every time.

Fully auditable. Rather than letting a model answer from memory, HMS built AskBarry so that every response is generated by running a live, traceable query against the governed data layer, with the query and source returned alongside each number. Users can see exactly where an answer came from and verify it for themselves. The difference between a tool finance teams can trust and one they can't.

Secure bydesign. The entire solution runs inside the firm's own cloud tenant. Data never leaves the environment and is never used to train a model. Role- and row-level access controls mean each user only sees whatthey're entitled to, and any action that writes back to a system of record keeps a human in the loop.

The Solution

With AskBarry in production, the experience is simple: a user types a question the way they'd ask a colleague, "Which of my relationships increased allocations last quarter?", and gets back a direct answer, the supporting table, and a chart, each tied to an auditable query. Because AskBarry connects to Salesforce through a headless integration, users can act on what they find without switching tools, updating records or triggering workflows right from the conversation.

What used to require a ticket and a wait is now a question and an answer. The data team has shifted from fielding one-off report requests to governing the foundation that makes every answer reliable.

Results

• Time to answer a typical data question reduced from hours or days, to minutes
• 99% reduction in ad-hoc report requests to the central data team.
• Every answer is traceable. Each response runs against a governed, auditable query the user can verify, replacing guesswork with a checkable source.
• Two months from kickoff to production go-live.

Quote

"In financial services, an answer you can't trace is an answer you can't use. Barry doesn't guess. It runs live queries against governed data, and every answer shows its work. We built it on Claude because the team needed to verify numbers, not just trust them, and that's exactly what we got. The question was never whether AI could analyze our data. It was whether we could stand behind the answers. Barry returns the query and the source for every number, chart, and dashboard. No black boxes, no queues, no waiting. (Well, very little waiting.) That's what it means to put AI in front of institutional teams." - Dave Masri, Founder and CEO, Gluon Digital

Why Claude

HMS built AskBarry on Claude Sonnet 4.6 because the firm's use case demanded a model that could reason carefully over complex, governed enterprise data and use tools reliably, not just generate fluent text. Claude connects to the governed data layer and to Salesforce through the Model Context Protocol (MCP), letting AskBarry turn natural-language questions into precise, auditable queries and take action on the results. Claude's strengths in instruction-following and safety also fit an environment where accuracy and data protection are non-negotiable.

About High Meadow Solutions

High Meadow Solutions is a data, AI, and Salesforce consultancy serving the financial services industry, with deep specialization in wealth and asset management. As a Services Partner in Anthropic’s Claude Partner Network, HMS helps firms build secure, auditable AI solutions on Claude, turning fragmented data into trustworthy, self-service intelligence. AskBarry is HMS's productized AI data analyst, designed to give every team member a direct line to their own data.

To learn how AskBarry could work for your firm, contact High Meadow Solutions