Santiago Cañas Casco

Built event-driven ELT and billing data systems that support 1B+ transactions/year across analytics and finance reporting. - Built a BigQuery, Pub/Sub, and GCS framework for historical replay, schema evolution, and long-term retention of proto events. - Designed dbt models and Prefect orchestration to make real-time warehouse transformations observable, easier to rerun, and more reliable. - Automated the client billing pipeline from warehouse aggregation to validated invoice drafts, applying contract rules dynamically and reducing manual finance preparation.

Master Class Collaboration

Reduced manual Jira data preparation time by 80% and gave product and analytics teams real-time visibility into sprint progress, capacity planning, and delivery bottlenecks. - Built Python extraction from the Jira REST API for issues, epics, sprints, story points, and users, with incremental warehouse ingestion. - Modeled sprint velocity, backlog evolution, and workload KPIs in dbt using staging, intermediate, and marts layers. - Orchestrated extraction, transformation, and data-quality workflows in Airflow with task-level alerts and retries. - Published curated Looker dashboards and custom reports that increased trust in Jira performance metrics across departments.

Reduced fraudulent transactions by 15% by analyzing transactional patterns with Python and PostgreSQL and surfacing anomalies for business teams. - Protected finance and leadership reporting by validating invoice and billing data before it reached Tableau dashboards. - Built Python data-quality checks for mismatched totals, missing client records, and inconsistent contract rules across billing pipelines. - Integrated validation into reporting workflows so Tableau dashboards reflected trusted financial data for revenue tracking and strategic decisions.

Improved pricing strategy visibility by linking mobile app sessions to transaction and pricing data in the DWH through a unique session_id model. - Built Python API ingestion pipelines and warehouse models for BI, pricing analytics, and operational reporting. - Enriched mobile event data with DWH metrics, enabling journey-level analysis of price offer acceptance across user flows, devices, and regions. - Created a journey characteristics heat map that helped product and marketing teams identify where users accepted or rejected price offers. - Used StreamSets to automate ETL transformations and validation checks from mobile APIs to warehouse datasets.

In order to satisfy the continually increasing demand for more capacity in fiber-optical networks, spectrally efficient modulation formats such as quadrature amplitude modulation with high cardinality have to be employed. However, non-ideal characteristics of optical transmitter components such as digital-to-analog converter, modulator driver amplifier and electro-optic modulator cause distortions of the generated optical waveforms leading to a reduced quality of transmission. One promising option to compensate these distortions are digital nonlinear pre-distortion filters applied in the transmitter digital signal processing unit. The goal of the thesis is to develop advanced methods and algorithms for transmitter impairment estimation in order to determine the optimum pre-distortion filters. The algorithms should be implemented and verified in Matlab. Finally, the implemented algorithms should be validated in laboratory experiments.

Escuela de Liderazgo Blas de Lezo consiste en una iniciativa llevada por alumnos con competencias en diversas areas desde marketing, derecho e ingeniería. Impartimos cursos de forma gratuita en distintos colegios e institutos de Madrid. La temática de los cursos es amplia, desde oratoria, gestión del tiempo, técnicas de estudio, dirección de equipos hasta gestión de redes sociales. El contenido de los cursos es teórico/práctico

We are designing a tool, which will enable us to measure brain signals. We test it in rabbits before we try in human brains. The project consists in two parts, one part is about programming software in order to handle the different signals we acquire and the other part is about electronics, we design an LNA ( Low Noise Amplifier)