Everything we build is a data system. Our research runs in two directions — making data systems work for AI, and inventing data systems for quantum computing.
We bring machine learning and data lakes together: integrating scattered, heterogeneous data into ML-ready training sets, discovering and selecting models from large model zoos, serving large language models straight from a database, and generating synthetic data across organisational silos without sharing raw records.
Model Lake
Amalur explores the convergence of data integration and machine learning — automating how scattered training data across silos is integrated for downstream models. It is the foundation of the group’s Model Lake vision, where heterogeneous data and rich model zoos meet in one place.
Synthetic data
SiloFuse generates cross-silo synthetic tabular data using latent diffusion models, so organisations can share realistic data without ever exposing raw, feature-partitioned records.
Time series
WaveStitch performs flexible and fast conditional time-series generation with diffusion models, stitching together realistic signals under user-specified constraints.
LLM serving
TranSQL serves large language models with relational queries — compiling model inference to SQL so that LLMs can run inside a database engine, even on low-resource hardware.
We reinvent data management for the quantum era — simulating quantum circuits inside a relational database, compiling database queries to quantum primitives, and charting how data should be stored, queried, and managed on noisy intermediate-scale (NISQ) quantum processors.