Carnegie adviser uses Claude for 401(k) match projections

Wendy Eldridge, a retirement plan adviser at Carnegie Investment Counsel in Ohio, uses Anthropic's Claude most days.

She most recently built match-formula projections for sponsors weighing a plan change. She described the workflow to 401(k) Specialist in the third installment of its Deep Dive series on AI in retirement planning, published Nov. 5. The projections cover the arithmetic a sponsor needs before a match decision: what different matching formulas would pay, for plans either changing the current match or moving to a safe harbor design. Running them through a model shortens the turnaround, which speeds up the sponsor's decision, Eldridge said. She also uses AI to condense SECURE 2.0 provisions for sponsors and to tighten the grammar and flow of fiduciary meeting minutes. "AI will absolutely not replace my experience and understanding of advising plan sponsors and individuals," she said.

For advisers running the same playbook, the practical gain is time in the sponsor's decision cycle rather than a new product to sell. A match comparison can reach a sponsor while the question is live instead of queuing behind client meetings, and a plain-language SECURE 2.0 summary can travel with it. Getting that far requires plan data privacy, review of model output, and human judgment kept in the advice. The record-keeping loop is tight, because the fiduciary minutes Eldridge puts through a model to fix grammar and flow are themselves part of the plan committee's record.

Eldridge said she is now working on applying the same tools to her business's back-office processes.