Training Turk

Asset & Wealth Management Expert

$100–130/hr · Mercor · Part time, 40 hours a week

Review and critique AI-generated investment recommendations by comparing them against real portfolio mandates and fiduciary standards.

What you would do

  • Grade model-built portfolio allocations against actual client mandates and constraints
  • Evaluate whether tax strategies proposed by AI systems hold up in real wealth management scenarios
  • Assess factor exposures and benchmark alignment in AI recommendations
  • Flag where AI reasoning is defensible on paper but breaks real client obligations
  • Document and explain assessment of AI reasoning quality in detailed written reports

Who they want

  • 5+ years hands-on portfolio management, research, allocation work, or wealth advisory experience
  • Deep familiarity with real mandates: clients, liquidity constraints, concentration limits, time horizons
  • Background in asset management, hedge funds, pensions, endowments, or private client advisory preferred
  • Clear written communication to document technical reasoning for technical audiences
  • Comfort with ambiguous tasks and ability to flag unclear instructions in project scope

Main skills

Portfolio constructionAsset allocationTax strategy

What the interview asks about

  1. 1.Mandate constraint violation detection

    AI systems often optimize mathematically without grasping real client restrictions; you must catch where recommendations breach agreements.

    For example: “An AI suggests 35% alternatives in a pension fund portfolio with a 15% liquidity constraint and required quarterly distributions. Explain why this recommendation fails and what allocation respects the mandate.”

  2. 2.Tax efficiency evaluation

    Knowing whether AI understands capital gains sequencing and tax-loss harvesting constraints in the client's actual situation shows analytical depth.

    For example: “An AI recommends liquidating appreciated tech positions to rebalance a taxable account client to a 60-40 target. What questions would you ask about the client's tax situation before accepting this choice?”

  3. 3.Factor exposure interpretation

    AI models may hide factor bets inside allocations without making them explicit; you must recognize when this matches or violates client expectations.

    For example: “An AI portfolio shows equal-weight small-cap tilts across three positions representing value, growth, and momentum. Is this aligned with a 'core long-only' mandate that restricts factor bets?”

  4. 4.Defensibility gap analysis

    Sometimes recommendations look reasonable statistically but conflict with real client economics; this tests whether you spot that mismatch.

    For example: “An AI recommends a 3-year drawdown strategy for a 65-year-old retiring today with a 30-year horizon and $2M to deploy. How would you assess the gap between paper defensibility and real longevity risk?”

  5. 5.Liquidity constraint reasoning

    Understanding how liquidity needs reshape allocation choices shows you grasp fiduciary reality beyond generic optimization.

    For example: “A high-net-worth client needs $500K annually from a $10M portfolio. How would you assess whether an AI allocation to 20% illiquid alternatives fits this requirement?”

  6. 6.Written assessment precision

    Most grading work happens asynchronously in writing; you must communicate judgment clearly enough that stakeholders trust your rating.

    For example: “Write a one-paragraph assessment of whether a 50/50 stock-bond allocation for a 25-year-old is 'defensible' despite being non-standard, explaining the mandate constraints you'd check.”

A task you may get

Assess an AI portfolio recommendation for a $5M household with growth-income goals, $300K annual spending, and 25-year horizon. Evaluate whether to accept or reject, noting key mandate constraints.

How to prepare

  • Review two real wealth management proposals you've worked on or know well, noting key constraint details
  • Practice writing a concise assessment distinguishing between 'mathematically sound but client-unsuitable' and 'genuinely good fit' allocations
  • Prepare a case where you've rejected a recommendation that looked good initially, explaining what constraint you found
  • Study how factor exposures get hidden in multi-asset allocations

The facts

Pay
$100–130/hr
Hours
Part time, 40 hours a week
Where
Remote
Field
Finance
Role type
Talent network
Posted
9/15/2026

We wrote this page from the public Mercor listing. It may be out of date, so read the full posting before you apply.