Quantforge started in 2015 with one question: can machine learning make portfolio cluster analysis accessible to people who aren't already quants?
The platform covers the full stack - from understanding how backtesting works to building ai trading bots that run on live data. Courses are structured around real scenarios: analyzing asset correlations, segmenting portfolios with unsupervised learning, and evaluating whether an auto trading bot strategy holds up under different market conditions. Students work with actual datasets, not toy examples. Group sessions run with a maximum of 12 participants so instructors can give meaningful feedback, and individual tracks let learners move at their own pace without skipping the hard parts.
Every instructor on the platform has built and run quantitative systems outside the classroom - crypto automation pipelines, binance trading integrations, bots for trading on live markets. Teaching is a second discipline, not a first job. That gap between theory and practice is something the team takes seriously, because most learners arrive having read the theory and hit a wall when they try to apply it.
Spent 6 years building clustering pipelines for institutional portfolio managers before joining Quantforge. Specializes in making backtesting results interpretable - not just technically correct.
Designed and deployed crypto bot systems on binance website infrastructure for 4 years. Teaches the engineering side - how auto trading bot logic actually executes under real latency conditions.
Designs the learning sequence from first principles to production-ready models. Her approach: a learner should be able to explain every decision in their cluster analysis, not just run the code.
