Python Alert System for Trend Detection
Time series
Built a lightweight alert engine that flags trend shifts using time-series indicators and configurable thresholds.
Designed for fast iteration and clean, human-readable output.
- Inputs: OHLCV/time-series data
- Logic: indicator thresholds + confirmation rules
- Output: clear alerts (what happened, why, and what changed)
Rule-Based Strategy Simulation
Evaluation
Simulated rule-based strategies to evaluate signal performance across market regimes.
Focused on disciplined testing and simple metrics that translate to decision-making.
- Compared signals across different volatility/trend environments
- Tracked hit-rate, average move, and false-positive behavior
Public Dataset Risk Analysis
EDA
Analyzed large public datasets to assess disruption risk with exploratory data analysis, basic statistics,
and stakeholder-friendly summaries.
- Cleaned and structured multi-column datasets for analysis
- Produced concise findings emphasizing risk-relevant patterns