Matteo Brucato

Helping humans make better decisions.

I build database-native systems for prescriptive analytics.
Decision-making, as declarative as SQL.

2025

Co-Founder & Chief Scientific Officer at OSM

2021

Researcher at Microsoft Research

Data Systems Group
MLO (Machine Learning & Optimization) Group
Azure VM Allocator

2021

PhD, Computer Science at UMass Amherst

DREAM Lab · advised by Peter J. Haas, Alexandra Meliou, and Azza Abouzied (NYU)

2017

Research Intern at Microsoft Research

Data Systems Group

2016

Research Intern at IBM Watson

2013

MSc, Computer Science at University of Bologna

2011

BSc, Computer Science at University of Bologna

Research

Publications →

My research on database-native prescriptive analytics spans six aspects — from query languages and execution engines to learning user intent, self-tuning, temporal reasoning, and fairness.

Videos

VLDB 2020 — Best Demonstration Award

SIGMOD 2020 — Technical Paper Talk

Talks

2024 Enabling In-Database Decision Support with Package Queries
Gray Systems Lab, Microsoft
2024 Wred: Workload Reduction for Scalable Index Tuning
SIGMOD 2024, Santiago, Chile
2021 Package Queries: Scalable Prescriptive Analytics Close to the Data
Georgia Tech; Carnegie Mellon University
2021 Stochastic Package Queries for In-Database Prescriptive Analytics
Sisu Data
2020–21 Package Queries: Enabling Declarative and Scalable Prescriptive Analytics in Relational Data
Harvard, DASLab; Megagon Labs; MIT CSAIL, Data Systems Group
2020 sPaQLTooLs: A Stochastic Package Query Interface for Scalable Constrained Optimization
VLDB 2020
2020 Stochastic Package Queries in Probabilistic Databases
SIGMOD 2020
2020 Stochastic Package Queries for In-Database Constrained Optimization Under Uncertainty
NEDB, MIT, Cambridge, MA; INFORMS 2020
2017 Scalable Package Queries in Relational Database Systems
Microsoft Research, Redmond; VLDB 2016, New Delhi
2016 Conceptual Queries for Multiple Data Sources
IBM Research, Austin, TX
2015 Package Queries: From Tuples to Packages
Berkeley Lab; AMPLab, UC Berkeley
2014 Improving Package Recommendations through Query Relaxation
VLDB 2014, Hangzhou
2013–14 Metric Spaces for Temporal Information Retrieval
ECIR 2014, Amsterdam; U. Bologna; Aarhus University

Service

Program Committee

ICDE 2027, VLDB 2027, SIGMOD 2026, SIGMOD 2025, VLDB 2025, DataPlat @ ICDE 2024, TKDE 2020

Guest Editorial Board

Information Processing & Management, special issue on Time & Information Retrieval, 2014/15

External Review Committee

VLDB 2017, SIGMOD 2015, VLDB 2015, RANLP 2013