$84/hr · Mercor · Hourly, 40 hours a week
Expert scientist generating and evaluating scientific training data for AI models learning semiconductor and thin-film processes.
What you would do
- Assess AI-generated technical reasoning and catch materials-science errors
- Devise difficult, high-level questions targeting advanced deposition and semiconductor topics
- Organize process parameters, characterization results, and technical knowledge into structured datasets
- Evaluate and rank model outputs according to established scientific precision criteria
- Design comprehensive training and evaluation rubrics for deposition systems
Who they want
- Hands-on ALD process development and optimization experience
- Extensive familiarity with coating and substrate processes in semiconductor and advanced-packaging contexts
- Solid understanding of molecular precursor chemistry and interfacial mechanisms
- Materials characterization experience (XRD, SEM, TEM, XPS, ellipsometry)
- Advanced degree (PhD/MS) in related fields: materials science, chemistry, chemical engineering, physics
Main skills
What the interview asks about
1.Process parameter optimization
Diagnosing what went wrong in deposition runs requires understanding interactions between precursor chemistry, chamber temperature, and surface kinetics.
For example: “You notice an ALD cycle produced rougher films than expected. Walk through how you'd adjust precursor pulse timing, purge intervals, or substrate temperature to investigate root causes.”
2.Evaluating AI model errors
The model must learn why certain parameter combinations fail; you must spot logical flaws in its explanations of deposition chemistry.
For example: “The AI claims increasing chamber pressure will improve film uniformity. Based on mass-transport principles, explain whether that's sound reasoning or where it goes wrong.”
3.Characterization data interpretation
Raw data from XRD, SEM, or spectroscopy often reveals subtle process problems; your ability to read and explain results trains models on real-world diagnosis.
For example: “XRD shows unexpected peak broadening and a small secondary phase. What does that tell you about film crystallinity and process fidelity, and how would you present those findings to a non-specialist?”
4.Structuring tacit knowledge
Models learn by example; translating your intuitive know-how into explicit, labeled datasets is how they generalize to new scenarios.
For example: “Describe how you'd document a successful ALD run. What parameters, intermediate checks, and acceptance criteria would you record so an AI system could learn the pattern?”
5.Technical writing precision
The model reads your written reasoning; ambiguous or imprecise language muddies what it learns about causation and trade-offs in materials science.
For example: “Explain in two sentences why XPS peak fitting matters for assessing precursor incorporation. Assume the reader has a chemistry background but not materials-processing experience.”
A task you may get
Given three published ALD parameter sets and their resulting film properties, design an evaluation rubric that scores on uniformity, purity, and process robustness.
How to prepare
- Review recent literature on ALD recipe design and chamber engineering to refresh process variables
- Prepare concrete examples from your past projects showing how you diagnosed and fixed deposition anomalies
- Summarize your hands-on experience with one or two characterization techniques in detail
- Write out your understanding of precursor chemistry and surface reaction sequence for a specific deposition system
The facts
- Pay
- $84/hr
- Hours
- Hourly, 40 hours a week
- Where
- Remote
- Open to
- USA
- Field
- Life, Physical, and Social Science
- Posted
- 8/1/2026
- Places left
- 9
We wrote this page from the public Mercor listing. It may be out of date, so read the full posting before you apply.