World models
We build AI mathematical models that learn the dynamics of markets, so portfolios can be optimized rather than guessed.
Quantitative finance · AI research
AI world models for
long-term portfolio returns.
MBot Labs builds AI mathematical models — grounded in cutting-edge research — to improve portfolio returns over the long term, the mid term, and on higher-conviction trades.
Approach
Research-driven, quant at the core.
Frontier research
We read and rigorously evaluate high-end research papers, distilling what holds up into models we can trade.
Three horizons
Long-term and mid-term portfolio optimization, plus a smaller, higher-risk allocation for higher-conviction trades.
Strategies
The quant toolkit we build on.
-
01
Momentum & reversal
Capturing persistent trends and mean-reverting moves across assets.
-
02
Statistical arbitrage
Exploiting relative mispricings between related instruments.
-
03
Execution & speed
Modeling timing and market microstructure to act on signals efficiently.
-
04
Transformers
Sequence models that learn complex, non-linear structure in market data.
Markets
Focused on stocks and crypto.
We concentrate on the equity and crypto markets for portfolio optimization — the domains where our models and data are strongest. We do not focus on commodities or oil.
Stocks
Crypto
Commodities
Oil
Founder
Vishal Khialani
Data scientist with a track record across financial and crypto projects — predicting prices and improving portfolios. At MBot Labs he leads research on AI world models for portfolio returns.
Reach out on LinkedIn →