Engineering Expert - Remote AI Fellowship

What You'll Do8

• Build prompts that test how well an LLM explains core concepts in hardware engineering and its subfields.
• Review model answers for accuracy, reasoning, and domain fit, then record where the model misses key details.
• Research engineering topics tied to your expertise, using AI tools to collect references and context.
• Label or rank model outputs to create training signals for partner AI labs.
• Work through assigned projects on an asynchronous schedule and meet the agreed 10 to 20 hours per week.
• Follow project rules for data quality, privacy, and confidentiality.
• Coordinate with AI lab partners through written updates and shared project spaces.
• Track advances in your engineering subdomain to shape new prompt ideas.

Requirements7

• Master's, doctoral, or postdoctoral standing in engineering or a related field. Current enrollment in a master's or PhD program counts.
• U.S.-based location and valid work or training authorization, such as F-1/OPT, J-1, or H-1B.
• Candidates on STEM OPT who need an i-983 cannot join. Approved i-983, pre-grad OPT, CPT, J-1, and H-1B situations are unaffected.
• Comfort with independent, asynchronous project work alongside AI labs.
• Expertise that can outmatch current AI systems in explaining important concepts in your engineering field.
• Availability of about 10 to 20 hours per week.
• Strong writing skills for prompt creation and response review.

Who Should Apply

An ideal candidate holds a master's, PhD, or postdoc in hardware engineering or a related engineering area, lives in the U.S., and has work authorization that does not require an i-983. The role suits engineers who can give 10 to 20 hours each week to asynchronous projects and who want to test their expertise against modern LLMs. The role is less suitable for anyone seeking a permanent full-time position with a fixed project pipeline or for engineers outside the U.S. who lack work or training authorization. A common reason applications score low is a vague description of domain knowledge. Another is an unresolved i-983 requirement, since Handshake cannot accommodate that case at this time.

Salary Insight