← All speakers

Bio, Work & Ideas

Jamie Neuwirth

Conference affiliation: Anthropic · 2024

Jamie Neuwirth is Anthropic’s Head of Startup Sales, helping founders translate Claude’s capabilities into viable products and scalable businesses. His career spans classroom teaching, education technology, payments infrastructure, and artificial intelligence.

Neuwirth taught math in the Arkansas–Mississippi Delta through Teach for America before becoming a program manager at Google, where he worked with school districts. By 2017, he was helping Stripe expand in Canada and supporting startups. After more than six years there, he joined Anthropic in 2023 to work on startup and midmarket sales. He attended Johns Hopkins University.

At Anthropic, Neuwirth approaches startup adoption of Claude through practical support: higher API rate limits, access to technical specialists, and guidance for founders developing AI-first product roadmaps. He envisions AI as a virtual cofounder, helping entrepreneurs prototype products and undertake work previously requiring specialized employees; he has identified healthcare as one promising opportunity.

At AI Engineer World’s Fair 2024, Neuwirth highlighted Claude 3.5 Sonnet and Claude Artifacts, emphasizing practical outputs such as code, diagrams, and scientific visualizations. Fellow Anthropic presenter Zack Witten led the subsequent prompt-engineering workshop.

Read the topics behind these talks

1 conference talk

Key ideas

Scroll to read ↓

A live workshop moves from an overly talkative patient agent to JSON extraction, writing examples, persona control, image inputs, translation grading, and grounded summaries.

  • What should a conversational agent say?
    0:35 ↗
  • Establish a baseline before cleaning up the prompt
    5:36 ↗
  • Replace vague requirements with observable behavior
    13:22 ↗
  • Reproduce the JSON failure—and notice truncation
    20:46 ↗
  • Control where generation starts and stops
    26:48 ↗
  • Check format in code; reduce oversized inputs
    36:26 ↗
  • Show the writing style you want
    42:17 ↗
  • Put useful preparation before the answer
    54:16 ↗
  • Separate persona routing from staying in character
    58:43 ↗
  • Role placement, counterexamples, and steering
    1:06:32 ↗
  • Improve the evidence in an image
    1:11:03 ↗
  • Calibrate what a grade means
    1:18:23 ↗
  • Analysis before probabilities
    1:27:14 ↗
  • Extract evidence before summarizing
    1:30:36 ↗

References