In their first three weeks on the product, each user had around 20 to 40 agents working for them. The most exciting use cases we've seen are:
- Truly live dashboards: agents own the data pulls, analysis and alerting. One PM described a live view of the short book in plain English and had it running in 30 minutes, refreshing every minute. It alerts the PM and other coverage agents when a threshold is hit or news breaks.
- Agents working as a team: agents build other agents, keep their own memory and run on Claude or OpenAI, so one model can check another. A memo gets separate bull, bear and auditor agents that catch errors and holes before the PM sees a number.
- Ideas surfaced without being asked: agents work overnight and in the background, built on the firm's own edge and framework. A Multiplier screen led to a deep dive that surfaced the key risks against the investor's framework, and it became an impactful position in the book.
Multiplier has also had a big impact on back-office operations. Three highlights:
- Bespoke investor materials: custom answers to allocator requests, drawn from the firm's own memos, write-ups, track record and past letters, turned around in hours.
- CRM management: an emerging manager's pipeline, outreach and follow-ups kept current by agents.
- NAV reconciliations: daily checks across funds, with a second agent verifying anything flagged.
Our clients are pushing the technology hard and unlocking more value as time goes on.
Claude vs Multiplier
Multiplier is a premium product designed from the ground up for hedge fund teams. Every affordance of frontier coding agents, but optimized for investors.
| Feature | Claude | Multiplier |
|---|---|---|
| Only Multiplier | ||
| Agent firstAgent-based, not chat-based. Never lose important information buried in chats you forgot you had. | ✕ | ✓ |
| Agents your whole team sharesShare one agent with the team: same chat, memory, files and schedules. Agents can message each other, so yours can ask a colleague's agent what it found. If someone else builds it, you can use it together. | ✕ | ✓ |
| Dashboards that do the researchAsk for a dashboard and the agent builds it, writes the code that refreshes it, and runs it on its own computer, with no repo or cloud environment to set up. It does the research and writes up what changed. Every refresh saves a snapshot of the numbers, and the team shares the agent that keeps it. | ✕ | ✓ |
| Fast exact search of connected SharePoint librariesWe index the text of Word, Excel, PowerPoint, PDF and email files, so agents find exact figures in seconds and open files in milliseconds. | ✕ | ✓ |
| Compute that scales to your workClaude serves millions of users, so each chat gets a fresh machine of a fixed size. Multiplier serves a small set of funds: every agent gets its own persistent computer, and we scale up compute resources based on investors' needs. | ✕ | ✓ |
| Models you chooseSwitch between Claude and OpenAI models, keeping memory and skills. | ✕ | ✓ |
| Spend you can traceSee spend by agent, by person and by routine, so you can measure the return on each workflow. The platform can also run in your own Azure. | ✕ | ✓ |
| Investing data built inPrices, financials, estimates, transcripts and filings, plus podcasts and X, ready in every agent. | ✕ | ✓ |
| Both | ||
| Coding agentsBoth write and run code to do the work, and both get better with every new model. | ✓ | ✓ |
| Your data licensesBoth use the data you already license. | ✓ | ✓ |
| Office filesBoth build Excel models, Word memos, PowerPoint decks and PDFs, and work in the models you already have. | ✓ | ✓ |
| Your toolsSharePoint, Outlook, Teams, Excel and your desktop. | ✓ | ✓ |
| Enterprise basicsSSO and spend limits. | ✓ | ✓ |