Simone Systems Research
Independent AI Systems Research
Research on how AI systems coordinate, verify results, allocate compute, and improve reliably.
Current work focuses on agent orchestration, evaluation, verification, compute economics, and adaptive AI systems.
Research Focus
Core questions guiding our systems development and empirical investigations.
Agent Orchestration
How should heterogeneous models, tools, and agents divide work and coordinate reliably?
Evaluation & Verification
How do we distinguish genuine improvement from false progress, benchmark noise, and brittle behavior?
Compute Economics
When does additional inference improve verified outcomes enough to justify its cost?
Adaptive AI Systems
How can systems learn from measured failures and improve their own workflows without losing control or reproducibility?
Research Projects
Prototypes, evaluation suites, and experimental systems implementations.
The Fleet
ActiveHow can heterogeneous agent topologies coordinate deterministically under strict latency and tool constraints?
Evidence-Gated Evaluation
ActiveWhat evaluation methodologies prevent benchmark overfitting and reliably detect subtle regression in multi-turn agents?
Cross-Model Verification
ExperimentalCan asymmetric verifier models provide formal safety bounds for black-box generative agents?
Pokémon Policy Evolution
ActiveEmpirical study on measured state tracking, environmental feedback, and policy adaptation in discrete environments.
Research Principles
Standard operating commitments applied across all experiments, evaluations, and published artifacts.
Evidence before promotion
Improvements should survive measurement, not merely look plausible.
Independent verification
Positive results should be reproduced outside the context that generated them.
Compute must earn its cost
Expensive inference should be reserved for decisions where it materially improves verified outcomes.
Negative results are retained
Failed hypotheses are useful evidence and should not disappear from the record.
Artifacts matter
Claims should refer to the exact code, model, configuration, and environment actually evaluated.
Research Notes
Working papers, benchmark reports, and technical reflections.
Compute Economics in Agentic Inference
Optimizing token context compression and speculative routing for low-latency agent architectures.
Evaluating Multi-Turn Agents: Beyond Single-Turn Benchmarks
Methodologies for measuring error propagation, goal drift, and recovery rates in autonomous workflows.
Architectures for Deterministic Agent Orchestration
Exploring state machine topologies and sandbox boundaries for reliable multi-agent execution.