Founding Simulation Engineer

Overview

This role sits between ML and product engineering. You take a simulation idea and make it operational: designing execution flows, state transitions, orchestration, data structures and runtime systems so simulations behave consistently at scale.

The role

The Founding ML Engineer focuses on models, behavioral engines, calibration, inference and learning systems. The Founding Product Engineer focuses on customer-facing products, workflows, interfaces, white-labeling and enterprise features.

The Founding Simulation Engineer connects them. You turn simulation concepts into simulation infrastructure.

What you would own

• Reusable architectures for product interaction, behavioral journey, ad, messaging, multi-agent, pricing and enterprise simulations
• Executable simulation flows: environment setup, audience initialization, stimulus exposure, branching, stopping conditions, outputs and reproducibility
• Agent orchestration: state assignment, model calls, sequencing, context, interaction, memory and event recording
• State and environment modeling: stimuli, event propagation, timing, feedback loops, interaction rules and persistent history
• Reusable schemas for audiences, environments, tasks, constraints, behavioral parameters and output metrics
• Observability through event streams, traces, timelines, decision logs, failure diagnostics, latency and cost monitoring

Problems you might solve

• Turn a 20-agent research prototype into a standardized workflow that can run thousands of times
• Coordinate 500 simulated individuals without every agent needing to communicate with every other agent
• Handle abandonment, hesitation, retries, changing minds and social influence
• Scale agent runs through parallel execution, asynchronous processing, queues, batching, retries and worker systems
• Create reusable primitives, including stimulus, environment, audience, decision event, interaction, observation, state transition and outcome

Reproducibility

A major responsibility is making simulations inspectable rather than arbitrary. You will build around seeds, configuration, versioning, deterministic components, experiment tracking, logs, replayability and simulation snapshots.

Researchers and customers should be able to understand what changed between two simulation runs.

Technical profile

• Python and/or Rust
• Distributed systems, backend engineering and asynchronous programming
• Event-driven architectures, workflow engines, queues and state machines
• Databases, APIs, LLM orchestration and cloud infrastructure
• Simulation systems and agent systems
• Game simulation, multi-agent systems, robotics, event simulation, distributed computing or RL environments are especially interesting

What this is not

This is not primarily frontend development, prompt engineering, model training, individual customer features or isolated scripts. The role exists to prevent Aetherya’s simulation technology becoming a collection of disconnected prototypes.

Compensation & involvement

This is a founding-team, equity-compensated position with an indicative grant of approximately 1.0%, depending on ability, commitment, ownership, experience, long-term involvement and contribution to core infrastructure. Vesting and final legal documentation apply. There is currently no cash salary.

We expect substantial, consistent involvement, approximately 20+ hours per week with flexibility around individual circumstances.

What success looks like

• Reusable simulation primitives and standardized configuration
• Scalable agent orchestration and distributed execution
• Reproducible experiments and event-based infrastructure
• Simulation observability and lower simulation cost
• New simulation products assembled from powerful platform configurations rather than custom engineering projects