We design data and AI systems grounded in domain knowledge to meet organisations’ operational needs.
Data science
We use statistical methods and modelling to explain patterns in data and support decision-making.
AI agent engineering
We design AI agent systems to carry out tasks within organisational workflows, with a focus on reliable, useful behaviour.
Software engineering
We develop applications that integrate with existing workflows and adapt as organisations’ requirements change.
CensoLab provides a natural language interface for analysing Chile’s 2024 Population and Housing Census. It performs calculations, generates maps, builds multivariate typologies, retrieves predefined geographic reports and interprets the results.
BCCh Studio lets users query the Central Bank of Chile’s economic statistics in natural language. It supports time series search and comparison, statistical analysis, visualisation and forecasting while preserving the traceability of data and results.
We frame the problem in its operational context and identify the human and technical constraints on the solution.
We design and develop solutions to fit the problem, with close attention to implementation details.
We integrate domain knowledge, data, methods, tools and AI agents so the system’s behaviour can be observed, evaluated and improved.
We integrate systems into their operational environments, taking infrastructure, security, documentation and maintenance into account. Operational feedback then guides how the system evolves.
Our team combines backgrounds in social science, statistics and programming.