Key Takeaways

  • Abby Barlow ran Westwood Management as a sole investment professional for three years without hiring junior staff by deploying Claude as a dedicated virtual analyst.
  • Document review cycle times dropped from three-day outside legal turns down to a single minute on an agreement containing 185 tracked changes.
  • Claude eliminated a recurring 60-tab Excel modeling task for fund vintage peer comparisons by writing a bespoke benchmarking app in five minutes.
  • Accuracy in automated investment office workflows depends on institutional memory capture via Barlow's Enterprise Context Interviewing Method.

The Barlow's Enterprise Context Interviewing Method

Abby Barlow developed a repeatable system to ground LLMs in family office history, personal working habits, and organizational guardrails:

  • Step 1: Onboarding Interview: Instruct Claude to conduct a 60-minute diagnostic interview asking exhaustive questions regarding the organization, business history, individual working style, and family preferences.
  • Step 2: Master Context Document Creation: Compile the structured answers into a standardized enterprise-level context document that defines organizational background and operational norms.
  • Step 3: Universal Project Ingestion: Inject the master context document at the initiation of every new project prompt to ground the LLM's domain awareness before assigning tasks.
  • Step 4: Prompt Loop Verification: Mandate that the model evaluate its final output against the original conversation objective by prompting it to 'review the deliverable, triple-check all the numbers again, find your own errors.'

When This Works (and When It Doesn't)

Applied when establishing automated workflows across an investment office to ensure institutional memory and specific organizational constraints are maintained across different tasks.

This method excels when a solo allocator manages diverse tasks across asset classes, from public equity reviews to LPA legal markups. By forcing the model to lead the diagnostic interview in Step 1, Barlow extracts implicit preferences that an allocator would otherwise forget to include in standard system prompts. The prompt loop verification in Step 4 catches hallucinated math and missed qualitative points before deliverables reach investment committees or family principals.

The framework breaks down when an office faces high-stakes legal execution or novel fund terms requiring statutory liability. Claude can parse 185 tracked changes to identify material commercial points in seconds, but it cannot sign off on fiduciary exposure. If an allocator uses the master context document without updating it after major strategy pivots, the model defaults to outdated institutional assumptions.

Why It Matters

Barlow's operational structure signals a permanent shift in investment office cost structures and staffing models. Historically, single-family offices and boutique allocators faced a stark tradeoff: pay substantial management fees to hire a bullpen of junior analysts, or drown the CIO in operational busywork like manual redline reviews and 60-tab Excel benchmarking models.

By treating LLMs as internal software engineers and junior analysts, single-person investment teams can match the operational bandwidth of mid-sized institutions. This shifts the allocator's primary asset away from team size and toward prompt architecture and prompt verification. Capital allocators who master institutional context injection can keep overhead flat while expanding due diligence throughput across fund managers.