Kyle Dickinson

Co-Founder

Pittsburgh, Pennsylvania, United States0 mo experience

Key Highlights

Built predictive pipelines with 95%+ AUC.
Analyzed high-frequency data for crypto markets.
Grew a digital community of 80,000+ followers.

Stackforce AI infers this person is a Fintech Quantitative Researcher with expertise in machine learning and financial modeling.

Contact

Skills

Core Skills

Quantitative ResearchPredictive ModelingResearch Skills

Other Skills

PythonData EngineeringFinancial RiskStatistical ValidationDiscrete MathematicsGraph TheoryCombinatorial LogicLinear AlgebraLightGBMMathematical FinanceXGBoostScikit-LearnDifferential EquationsBERT (Language Model)FastF1

About

I’m a Mathematics student at CMU focused on the intersection of high frequency data, machine learning, and financial modeling. My work ranges from quantifying price discovery in crypto-asset markets to building 95%+ AUC predictive pipelines for alternative credit scoring. Beyond the technical side, I’ve built and scaled a digital community of 80,000+ followers on Instagram where I simplify complex financial and technological topics for a global audience.

Experience

0 mo

Total Experience

Average Tenure

Current Experience

Carnegie mellon university

Quantitative Researcher (Independent)

Nov 2025 – Jan 2026 · 2 mos

L2 Data Analysis: Analyzed high-frequency Level 2 tick data from Bybit using an event-based aggregation framework to normalize for non-linear information arrival.
Predictive Modeling: Proved that net signed trade imbalance drives price discovery, explaining up to 47.1% of price variance (R^2) across macro scales.
Statistical Validation: Conducted ANOVA and t-tests (p < 0.01) to identify a staircase effect in bid-ask spreads during high-volatility regimes, confirming adverse selection models.
Liquidity Fragility: Quantified structural asymmetry where sell-side price impact exceeds buy-side by $30.46 at the 100k-trade scale, providing empirical support for Margin Spiral theories.
Infrastructure: Developed Python pipelines to synchronize limit order book snapshots with trade feeds, filtering for API latency and crossed-book anomalies.

PythonData EngineeringFinancial RiskStatistical ValidationQuantitative ResearchPredictive Modeling

Louisiana state university

Student Researcher

Jun 2022 – Jul 2022 · 1 mo · Remote

Mapped subgraphs across 16 edge-color relations using Ramsey’s Theorem and monochromatic clique proofs, minimizing complexity to 9 non-isomorphic classes
Categorized complex graph architectures into three primary classes by modeling clique and bipartite interactions via isomorphic filtering

Discrete MathematicsGraph TheoryResearch Skills