New York equity hedge fund · $1bn–$5bn AUM · 25–50 staff
The firm was already using AI. The missing piece was a shared way to move forward.
One practical day brought investment, technology, operations and compliance together around real workflows, reusable skills and the decisions required to turn experimentation into a coordinated program.
What followed
A funded next step
Twelve days later, the firm commissioned a separate discovery phase to solve the research-data constraint the day had exposed.
- Built around the firm’s real work
- 1 day
- Learning and deciding together
- 4 functions
- Reusable workflow starting points
- 14 skills
- To commission the next phase
- 12 days
Capability varied widely. Some staff were learning how to give a model enough context to return a useful answer. Others had connected data sources, written code and assembled detailed research workflows. People were moving, but in different directions.
The issue reached beyond training. Investment professionals wanted better earnings preparation, monitoring and idea work. Technology had to assess access and integration. Compliance needed to understand where information travelled and where human review belonged. Operations needed examples grounded in its own responsibilities.
Altitude 7 designed the day around the firm’s research cycle and technology choices. The purpose was to give everyone the same map, then use it on work they already recognised.
The learning path
Start with a common language. Finish with a decision the firm can act on.
The agenda moved from concepts to role-specific practice, then connected the strongest workflows to their data, control and ownership requirements.
Firm-wide AI day
Learn together · practise by role · decide what to build
Shared foundation
Common language for models, tools, sources and review
Role-based practice
Investment, technology, operations and compliance
Live workflows
Earnings, monitoring, research and operating tasks
Reusable skills
Structured instructions with sources and safeguards
Next decision
Separate the ready-now work from infrastructure needs
Work brought into the room
Investment research · earnings preparation · thesis monitoring · operating workflows
Controls carried through
Visible sources · uncertainty stated · human review · predictable outputs
Practise on real work
The team did not need another list of AI tools.
Short instruction blocks were followed by live examples and hands-on work. Participants saw how role, task, context, output structure and examples change a model’s answer. They also learned how to ask a model to challenge an investment thesis, surface uncertainty and examine a company through different analytical lenses.
The workflows followed the firm’s own cycle: idea generation, pre-earnings preparation, results analysis, post-earnings follow-up, thesis monitoring and investment-committee work. Operations, investor relations, accounting and compliance received relevant examples from their own work.
The same principle ran throughout the day: an answer is only useful when the user can see what evidence supports it, what the model inferred and what still needs human judgment.
Reusable starting points
Show the difference between a prompt and a controlled workflow.
Altitude prepared a working analyst environment and fourteen reusable starter skills for the session. Each was a structured starting point: clear instructions, expected sources, a predictable output and safeguards against invented or unsupported information.
Investment research
Earnings preparation, thesis checks, deep research and investment memos
Daily intelligence
Morning briefs, monitoring and structured follow-up
Firm operations
Meeting notes, investor responses and variance analysis
A live earnings workflow demonstrated how a repeatable process can retrieve evidence, apply the firm’s analytical lens, expose uncertainty and return a structured brief for review. These were training assets and workflow examples, not unattended production agents.
Independent platform analysis
Choose from the workflow backwards.
The firm was evaluating specialist research platforms as well as flexible general models. Altitude assessed the options against the work: what information each could reach, how deeply it could retrieve it, what controls came with the platform and what the firm would still need to build.
That shifted the discussion away from choosing one headline tool. The team could instead ask what recurring outcome it wanted, which internal and external sources were required, where model judgment entered, who checked the result and which workflows needed shared infrastructure.
The comparison also made an important boundary visible. A polished demonstration may work with public data, while the valuable version depends on internal notes, email, files and licensed research that the tool cannot yet reach.
The evidenced result
The day exposed the next constraint—and gave the firm enough clarity to act.
The client described the day as a success and asked to schedule next steps. The strongest opportunities depended on research spread across internal notes, email, shared folders and external systems.
Twelve days later, the firm commissioned a separate data-discovery phase to test retrieval in a controlled environment before building a wider agent layer.
See what the discovery establishedThis engagement established a shared vocabulary, practical starting tools and a defined next problem. It did not measure firm-wide adoption or deploy production agents.
The service behind this work
AI Training & Fluency
Role-specific training, live-work clinics and power-user tracks that build practical capability across the firm.
See how AI training works →This engagement also touched Data & AI Foundation.
Read next
AI training
PDF reports into model-ready data
Three working sessions on specialist extraction, Excel model updates and agents the investment team could test.
AI training
Building agents the team could challenge
A Claude Code programme on agent scoping, sourced evidence, testing, multi-agent design and the decisions behind shared agents.
Is your team already experimenting, but without a shared way to scale what works?
Bring the people who own the work, the data, the technology and the controls into one practical session. We’ll build the day around the workflows your firm is already trying to improve.