Private Draft

The 29 personas behind AI

We’ve organized every stage and persona in the AI supply chain, informed by real recruiting at frontier companies. Click any row to see matching profiles from our talent graph.

Shaped by Industry Experts

← ATOMS & ENERGYUSERS & MARKETS →

← Back

Frontier Research

Authors SoTA research bets

Frontier Research

Summary

Known as: Research Scientist, Research Fellow, Research Engineer, Member of Technical Staff, Research Scientist (World Models)

Pushes the state of the art in foundation model architectures (language, vision, multimodal), scaling laws, and novel training techniques. Publishes papers that precede each generation of AI capability. High leverage, low headcount, and often heavily credentialed. At frontier labs, this also includes capability forecasting: predicting what models will be able to do at the next scale using empirical scaling relationships.

Specializations

Architecture & Scaling — Novel architectures (transformers, SSMs, mixture-of-experts), scaling laws, efficient training methods, and multimodal integration. The structural bets that define what the next generation of models looks like and how they scale.

Capability Research — Reasoning, code generation, tool use, long-context, multimodal understanding, and continual learning (online adaptation, persistent memory, learning from deployment experience) — pushing what models can do. Includes the experimental work that turns architectural ideas into measurable capability gains.

Understanding & Predicting Training — Empirical and theoretical work on training dynamics, scaling laws, generalization, loss landscape analysis, and capability forecasting. Predicts what models will be able to do at the next scale and informs resource allocation and safety planning.

A growing sub-population has deep domain expertise in biology, chemistry, or medicine — applying foundation model capabilities to drug discovery, protein structure, genomics, and materials science. Concentrated at dedicated labs (Isomorphic, Recursion, Arc Institute) and frontier lab science teams.

Where the Work Lives

[1]Substrate

[2]Compute

Secondary

Scaling laws research defines the relationship between compute investment and model capability, directly shaping infrastructure strategy.

[3]Intelligence

Primary

Publishes the architectural and scaling breakthroughs that define each generation of model capability.

[4]Systems

[5]Distribution

Candidate Archetypes