Mechanical Engineering Expert - AI Model Evaluation

Get paid to help build better AI.

Alpheva AI, the Silicon Valley company that is building intelligence behind frontier AI, is looking for experienced mechanical engineering professionals to participate in paid, project-based work that helps train and evaluate the next generation of AI systems.

You work remotely and asynchronously on projects that fit around your existing schedule.

Your mechanical engineering expertise becomes the signal that helps AI systems learn what accurate, rigorous, and professionally sound engineering work actually looks like.

IMPORTANT: To be considered, Apply directly through: alpheva{Dot}ai/experts/

What you'll do

Depending on the project, you may:

Analyze: Review engineering problems, designs, calculations, specifications, technical drawings, and mechanical systems.
Evaluate: Review AI-generated engineering solutions, calculations, designs, explanations, and reasoning.
Validate: Determine whether AI-generated engineering outputs are technically accurate, logically sound, and consistent with engineering principles.
Review: Assess mechanical designs, engineering calculations, simulations, material selections, system requirements, and technical documentation.
Reason: Solve realistic mechanical engineering problems and explain how an experienced engineer would approach them.
Compare solutions: Assess multiple AI-generated engineering solutions and identify meaningful differences in accuracy, efficiency, safety, reasoning, and professional judgment.
Create evaluations: Develop challenging mechanical engineering tasks and criteria for measuring AI systems.
Provide expert judgment: Identify errors, incorrect assumptions, design weaknesses, inconsistencies, and technical risks in AI-generated engineering work.
Projects may involve mechanical design, thermodynamics, fluid mechanics, heat transfer, materials, manufacturing, machine design, robotics, dynamics, control systems, CAD, simulation, and other areas depending on your expertise.

Who we're looking for

We're looking for experienced mechanical engineering professionals, not necessarily AI researchers.

You should have:

3+ years of professional experience in mechanical engineering or a related field
Strong engineering and analytical reasoning skills
Experience working with real-world engineering problems and systems
Ability to evaluate engineering analysis, designs, and conclusions critically
Strong attention to detail
Ability to identify errors, inconsistencies, incorrect assumptions, and technical limitations
Strong professional judgment
Ability to clearly explain why an engineering solution is correct or incorrect
Experience with any of the following is valuable:
Mechanical design and machine design
CAD and engineering drawings
Thermodynamics
Fluid mechanics
Heat transfer
Materials and material selection
Manufacturing and production
Mechanical systems and equipment
Robotics and automation
Dynamics and vibration
Engineering simulation and analysis
Product development
Engineering standards and specifications
PE, Chartered Engineer, or equivalent professional qualification
You do not need prior experience working in AI.
What matters most is that you know how to do mechanical engineering work well.

You might be asked to:

Review an AI-generated mechanical engineering solution and identify the issues.

Evaluate whether an engineering calculation is technically correct and identify incorrect assumptions.

Review a mechanical design and assess whether it satisfies the stated requirements.

Compare two AI-generated engineering solutions and determine which is stronger.

Evaluate an AI-generated analysis involving thermodynamics, fluid mechanics, materials, or mechanical systems.

Analyze a realistic engineering scenario and explain your reasoning.

Identify calculation errors, design weaknesses, inconsistencies, or unsupported conclusions.

Evaluate whether an AI-generated engineering recommendation is technically feasible.

Create challenging mechanical engineering problems that can be used to evaluate AI systems.

The goal is not simply to determine whether engineering work looks reasonable.

The goal is to capture the judgment an experienced mechanical engineer uses to determine whether engineering work actually makes sense.

Compensation

Paid project-based work.

Rates vary based on your experience, specialization, and project requirements. Your applicable rate and project scope will be shown before you commit to a project.

Some projects are short assignments. Others may involve recurring work over a longer period.

There is no fixed schedule and no minimum commitment.

Work arrangement

Remote
Asynchronous
Project-based
Flexible hours
Work around your existing job or commitments
Paid for qualifying project work
But producing a plausible-looking engineering answer is not the same as demonstrating strong engineering judgment.

AI needs to learn how experienced mechanical engineers:

Analyze engineering problems
Apply fundamental engineering principles
Evaluate calculations and assumptions
Assess mechanical designs
Select appropriate materials and methods
Evaluate safety, feasibility, and performance
Interpret technical specifications and requirements
Identify errors and inconsistencies
Recognize important edge cases
Compare engineering solutions
Explain technical conclusions clearly
Determine when engineering analysis is actually reliable
That's where you come in.

Your mechanical engineering expertise helps shape how AI learns to reason about engineering.

How it works

Apply
Tell us about your mechanical engineering background, experience, qualifications, and areas of expertise.

Get qualified
Complete a short assessment designed around the type of mechanical engineering work you'll perform.

Get matched
When your experience matches an active project, we'll share the scope, requirements, and compensation.

Do the work
Complete projects remotely and asynchronously.

Get paid
Receive the agreed rate for your completed work.

Help teach AI how great mechanical engineers actually work.

Apply directly through: alpheva{Dot}ai/experts/