Separate discovery from delivery
A useful brief explains whether the team needs to develop new methods, test existing approaches or build dependable products around them. Titles alone do not settle this. Describe what the person will deliver and the constraints they will face.
Research scientist
For a research-led role, specify the questions to investigate, the expected research contribution and how progress will be assessed. Relevant evidence might include publications, rigorous experiments or a clear explanation of failed approaches. Connect that evidence to the actual work.
Research engineer
Clarify how much of the role involves implementing experiments, improving training systems, building evaluation infrastructure or helping research reach production. A strong brief describes the relationship between scientists, engineers and product teams.
AI engineer
For applied AI work, explain the product, the available data, the evaluation requirements and the operational constraints. Discuss reliability, latency and cost in the context of the role. Assess the candidate’s decisions and delivered systems, rather than relying on tool names alone.
What this looks like in practice
Before Corellia Talent, Josh placed a Staff Engineer into a Series B AI startup in San Francisco and a researcher with an exceptional academic and competition record into a pre-seed startup. These were different briefs. The useful question is which evidence relates to the work this particular team needs done.
Read the placement accountCheck the compensation assumptions.
Our salary guide separates advertised pay, disclosed equity and bonus from planning estimates. Compare the role, location and company context before using a figure in a decision.
Explore the evidence and compensation plannerQuestions founders ask
What is the difference between an AI engineer and a research engineer?
An AI engineer often focuses on delivering AI capabilities in a product. A research engineer may build experiments, evaluation systems and tooling for research. Agree the expected outputs because titles overlap.
When do we need a research scientist?
A research scientist role is relevant when the work centres on new methods, hypotheses and experimental findings. Define the questions, resources and evaluation criteria before choosing the profile.
Do all AI roles require a PhD?
No universal qualification requirement follows from an AI job title. Set requirements against the actual work and assess relevant contributions.
Which roles does Corellia cover?
Corellia supports AI engineers, ML engineers, research engineers, research scientists and applied scientists, alongside software engineering and technical leadership.