AIVIA Partnerships for University Labs
For faculty: what companies are actually evaluating.
For researchers: a path to specialized roles.
STEM-Away has run technical virtual internships for thousands of participants. AIVIA Evaluations grew out of that work: scenario-based assessments companies use to see how candidates reason through realistic engineering problems. Several industry partners have also asked us to help them connect with researchers working in areas relevant to their open roles. This partnership creates that connection.
The Partner Lab Report
Partner labs receive a periodic report based on activity across AIVIA, including:
- which evaluations different types of companies are requesting
- which technical areas are seeing increased demand
- which failure modes companies are choosing to evaluate
- which reasoning dimensions industry partners are weighting most heavily
- where new demand is emerging across specialized roles
For example, an autonomous-driving report could show whether companies are placing greater emphasis on perception failures, planning and safety, hardware-aware deployment, real-time deadline misses, or quantization accuracy cliffs.
The report reflects aggregated AIVIA activity. It is not an assessment of individual researchers in your lab.
A Personalized Path for Your Researchers
We will map up to five researchers from your lab to the AIVIA Evaluations that best fit their experience and target roles, at no cost to the lab or researcher.
The mapping gives each person a focused starting point: which evaluations can help them demonstrate capabilities that may not be visible from a resume alone.
Researchers decide whether to complete the recommendations and whether to make their results visible to employers.
What Are AIVIA Evaluations?
AIVIA Evaluations are dynamic, personalized conversations built around realistic technical scenarios. Each evaluation focuses on a specific engineering component and its real-world failure modes, with follow-up questions that adapt to the candidate’s reasoning.
In autonomous driving, for example, our Hardware-Aware Model Mapping evaluation asks researchers to diagnose why a model that appears efficient in its framework does not reach the expected throughput on its target accelerator.
Evaluations are open-book and include:
- 3–5 questions with adaptive follow-ups
- detailed feedback and rubric-based scoring
- a full report with strengths, gaps, and question-level analysis
- a brief voice reflection for verification
What matters is the quality of the researcher’s reasoning, not what they have memorized.
What Researchers Gain
Researchers build a reusable, verifiable portfolio of how they reason through difficult technical problems: a GitHub for technical reasoning.
They receive:
- detailed evaluation reports
- rubric scores, strengths, and gaps
- verified resume tags and shareable proof artifacts (e.g., evaluation reports they can link to)
- career guidance based on their results
- optional visibility to hiring teams
Hiring teams can search AIVIA by domain, engineering component, evaluation score, resume tags, and reasoning patterns. Researchers control which results are visible and can opt in to company introductions when their completed evaluations align with an open role.
No Cost. Researcher Controlled.
There is no cost to the lab or participating researchers.
Researchers choose whether to complete an evaluation, make a result visible, or opt in to company introductions.
Schedule a conversation →
or email partners@stemaway.com for more information.
Ready to join? Email partners@stemaway.com with “Interested” in the subject line and we’ll send the next steps.