René Felipe Quezada Castañeda Email & Phone Number

Santiago, Cl

Tierra Amarilla, Región De Atacama, Chile

Principal & lead data scientist for Minera Candelaria, responsible for the Advanced Analytics area.- Leading the design and business case of the first comminution's integrated recommendations framework for Lundin Mining, based on process informed-prescriptive analytics.- Leading the development of new tools and resources, focused in reportability, with more than 10 new products, involving new business areas as clients of the Business Insight area and optimizing existing reports.- Leading the process of migration for key reports of the Candelaria’s Advanced Analytics area to the centralized corporate cloud computing environment, aligned with best practices and directions of the corporate Data & Analytics Direction and the Regional Information Technologies Department of Lundin Mining.- Lead and developed the baseline and business case definitions that act as the foundation of the main machine learning initiatives of Candelaria and, by extension, of Lundin Mining, based on tested projects already developed by the Advanced Analytics area under the Full Potential initiative, with validated value capture assessments.- Lead and coordinated the implementation of a data-driven operational standardization project in the grinding area of the Candelaria concentrate plant, whose target is the building of an operational recipe that converges with the throughput maximization of the mills with variability reduction. This project has a documented value capture assessment, already validated, currently being scoped as an initiative for other Lundin Mining's sites.

Tierra Amarilla, Región De Atacama, Chile

Senior business analytics engineer for the Business Insight Department of Candelaria. In charge of the development of data science projects with the target of monitoring, predicting and optimizing the operational conditions of the mining and metallurgical areas of Candelaria.- Took lead in the development and implementation of a data-driven operational standardization project for the crushing and grinding areas in the concentrate plant of Candelaria District, in the context of the Full Potential initiative currently being carried on. This project is currently under implementation via industrial testing.- Managed the standardization of analytics projects development for process optimization of all Candelaria business areas, mostly oriented to the gross copper maximization (after process). - Led the searching of third-party technological partners for cooperative developments regarding key insights for Candelaria’s operations, having developed the technical requirements definition for all services, including SaaS, collaborative development and engineering support, and the corresponding standards for analytics projects development.- Lead the process of standardization of the data engineering convention for the developing of advanced analytics-based digital products in Candelaria, including the definition of data dictionary formats and the structure of data science projects.- Lead the development of a modular and parameterized data infrastructure for the building of grinding and flotation predictive models in the Candelaria’s concentrate plant, with high scalability level in terms of the addition of parallel and downstream processes, and capacity for all kinds of analysis, including context-based ETL processes ranging from the capture of historical data in mining and processing areas to experiment tracking and model logging in deploylment frameworks (Kedro + MLFlow).

Tierra Amarilla, Región De Atacama, Chile

Business insight consultant for the operational excellence area of CCM Candelaria. In charge of the management of the data architecture and strategy to further develop digital applications to enhance the decision making of the Company on a data driven basis.- Led the development of a factor-wise grinding’s throughput model for further strategical loss management at a full operational and technical scale. The target of this model is to automate the explanations of peaks and falls of grinding throughput daily.- Development of business analytics methodologies to define the operational standard for the process control of the grinding and flotation plants.- Design and development of the data governance and infrastructure standardization for advanced analytics based projects leading the optimization and transformation of Candelaria.

Santiago, Región Metropolitana De Santiago, Chile

Assistant teacher for the mine-applied Data Analytics subject at Universidad del Desarrollo, Chile.- Design and development of participative classes.- Covering of all key areas in this subject. These include data structures, complete handling of this data (cleansing and wrangling) and reportability.- Pythonic focus, including classical and highly utilized libraries, such as Numpy, Pandas, Matplotlib, Plotly and Seaborn.

Chuquicamata, Antofagasta, Chile

Specialist data scientist & data analyst for the first ever Divisional Advanced Analytics Area of Codelco, in charge of the development of key initiatives to value capture in all key business areas of Codelco Chuquicamata. These include:- Complete characterization of ore sources based of kernel density estimations (KDEs) in order to identify the best ore suppliers from each origin feeding the Chuquicamata plant.- Key participation in an end to end mine to mill optimization project led by the Digital and Analytics Department of Codelco, in order to maximize the copper production. Operational member of specialist data scientists team, in charge of key improvements in the grinding and flotation machine learning and optimization models, and support for ore tracking modelling team.- Development of a loss distribution model for the collective flotation process to ensure a correct management of these losses regarding several key features, making it available as a dashboard report.

Calama, Antofagasta, Chile

Operational Excellence Lead, in charge of continous improvement of mine to mill KPIs, including the development of predictive models for several business areas (grinding, flotation, machine reliability and safety propension classification). Most notable achievements include:- Collapse-damage classification model for production level pillars in the Chuquicamata underground mine.- Reliability predictive model for the upholder ball-mill in the SAG grinding circuit.- Development of adherency follow-up strategy and reportability for the operational standards in all the business areas of the copper concentrate plant of Chuquicamata.- Key participation in an end to end mine to mill optimization project led by the Digital and Analytics Department of Codelco, in order to maximize the copper production. Member of specialist data scientists team, in charge of key improvements in the grinding and flotation machine learning and optimization models.

Chuquicamata, Antofagasta, Chile

Operational excellence expert, in charge of the development of machine learning models for conventional grinding processes in the Process Engineering Department of Chuquicamata's copper concentrate plant, as a part of an advanced analytics program dependant of the Operational Excellence Department. Key responsabilities include:- Advanced analytics applied to find value-capture opportunities in specific KPIs of the plant, mostly grinding and flotation.- Collaboration with an end to end, mine to mill, optimization project (as a data scientist and translator) to develop a fine copper recommender system for the concentrate plant, including in-situ support for its conception and deployment in productive environment.- Development of feature process mappings for all key areas of the concentrate plant, including secondary-tertiary crushing and all grinding and flotation circuits.

Codelco Chile, División El Teniente

Researched and developed predictive machine learning models for critical caving mining phenomena occurrence, including collapse processes in production sectors and incoming projects, and relevant induced seismicity and overbreak in tunnels. Worked on analysis of pre-conditioning radius of influence, in terms of the water propagation pressure, in hydraulic fractures induced in the rock mass by this technique for further cost savings in boreholes height. Development of an experimental deep learning model for the estimation of the occurrence probability of relevant seismic events on a month by month basis for incoming key projects. Principal achievements include:- Development of a pilot project of a deep learning model for the estimation of the density of radiated seismic energy, based on caving growing, extraction rates, hydraulic fracturing conditions, geological heterogeneity of the rock mass and stress conditions, for production sectors and incoming projects, on a year by year basis, taking the northern region of El Teniente mine as a training set. - Took part in the development of a machine learning model for the estimation of seismic response and overbreak in critical tunnels emplaced in high-stress environments, based on operational and geological features.

Codelco Chile, División El Teniente

Researched and developed studies and back-analysis for historical collapse events in El Teniente, encompassing the whole primary ore extraction begun with the Teniente 4 Sur operation, building a state-of-the-art of this phenomenon for the panel caving operations, with further application on the building of a machine learning classifier model for the estimation of collapse response in production level pillars, taking Esmeralda and Reservas Norte as a training set, inputting the stress and strain conditions in each production sector, induced seismicity, rock mass’ geo-fabric, operating and design parameters and the mining method’s architecture to the learning algorithm. Provided support engineering in medium term geomechanical planning and data-analysis for a variety of inquiries, including pre-conditioning, caving line geometries, caving rates and cave-back progression. Principal achievements include:- Took part in the development of a full-scale machine learning model for the estimation of collapse-damage probability in production level support pillars at El Teniente Mine.- This work was presented in two congresses: APCOM 2019: "Mining goes digital", and GEOMIN MINEPLANNING 2019.

Codelco Chile, División El Teniente

Researched and developed analytical models on collapse instabilities in production sectors of El Teniente mine. Provided support for conciliation analysis of operating parameters against design values for further geomechanical guidelines’ update. Principal achievements include:- Development of a first version of a full scale classification model for the estimation of a collapse-vulnerability index for production level support pillars in production sectors of El Teniente mine.- Development of the conciliation and summarization of all studies carried out by twenty eight years of primary ore caving mining research, in order to state an integral mining design model for the minimization of collapse occurence in high stress mining environments, with significant rock burst hazard.