Forecasting
Empirical studies of price dynamics, volatility, liquidity, and regime structure across financial and prediction-market datasets.
Research on uncertainty, markets, and model behavior
Kernel Research
Researching AI systems for financial and prediction markets. We examine how language models, agents, and adaptive learning systems form hypotheses, represent uncertainty, and behave in noisy information environments.
01 Forecasting
02 Simulation
03 Risk
04 Agents
Research areas
Kernel research program
Kernel research program
Models, agents, and market systems examined through testable hypotheses.
Decision models
Experimental simulations for agent policies, sequential decisions, and feedback loops under uncertainty, constraints, and incomplete information.
Control
Protocols for uncertainty estimates, exposure, counterfactuals, model drift, and failure modes before results are interpreted as robust.
Method
Research-driven by design
Research-driven by design
Each system starts as a research question and is tested through measurement.
Our work combines machine learning, quantitative research, market microstructure, and research software. We study how models learn from data, encode assumptions, and change behavior as information conditions shift.
The goal is not an isolated benchmark result. It is a reproducible process for formulating, testing, and documenting hypotheses with clear assumptions, limitations, and failure modes.
00Dataset definition and market context
01Hypothesis formation and literature mapping
02Backtesting, ablation, and stress scenarios
03Agent-policy evaluation under controlled constraints
04Out-of-sample evaluation and drift analysis
A research program for studying signal formation, uncertainty, and reflexivity in
markets shaped by human expectations.
Prediction markets
Research on event probabilities, information arrival, narrative formation, liquidity, and probabilistic belief updating.
Financial markets
Studies of price dynamics, volatility, regime shifts, market microstructure, and risk-aware evaluation.
LLM and agent infrastructure
Evaluation harnesses, tool-using agents, memory studies, and decision logs for auditable research artifacts.
Research notes
Selected directions
Selected directions
An index of current research questions, experiments, and systems under study.
Market-regime detection with adaptive models
LLM-assisted hypothesis generation in noisy information environments
Agent-policy evaluation for simulated decisions under uncertainty
Uncertainty quantification for autonomous market models
Benchmark design for agent reasoning under market uncertainty
Our team
Researchers and engineers studying AI systems under uncertainty.
Kernel Research brings together PhD researchers, quantitative scientists, machine learning specialists, and systems engineers for a focused research project on uncertain markets and agent behavior.
The team combines academic depth with implementation discipline: machine learning, high-performance computing, mathematics, and reproducible research infrastructure.
Febin Varghese
Quant LeadEx-McKinsey. MSc Mathematics, First-Class. British Maths Olympiad. Research strategy, mathematical framing, and institutional analysis.
Dr. David Liu
LLM / Agentic Systems ResearchPhD Cambridge in computational neuroscience and machine learning. #2 Part III Physics. International Physics Olympiad, Dutch team 2015. Ex G-Research and Meta AI.
Ronan Moynihan
System ArchitectSystems architect and full-stack engineer with experience building market data, execution, and research infrastructure from first principles.
Dr. Yurii Piadyk
Core Systems EngineerPhD NYU in Computer Science. High-performance computing and algorithms. Compilers, solvers, and numerical methods.
Jai Duhra
Lead Developer / Blockchain EngineeringFull-stack blockchain developer focused on smart contracts, DeFi integrations, and blockchain systems used in market and agent research.
Contact
Research inquiries
Research inquiries
Interested in uncertainty, markets, or agent research?
Send a short note on the research question, dataset, evaluation protocol, or methodological problem. Kernel Research is for academic and exploratory study; it does not provide investment advice, trading services, or live execution systems.