Javier Guzmán Figueira Domínguez

Additionally to the maintenance and improvement of previously developed services and tools, focused on building a solution that oversees the quality of speech data delivered to clients and eases the delivery process.

Among other details, this system includes the development of:

• a micro-service and a REST API (FastAPI),
• a binary classification model for audio (deployed in Kubernetes using MLflow),
• an on-demand metrics computation using Druid.

As well, cooperated with stakeholders to understand the impact of our solutions on their needs and guide them on its usage. And, helped new team members in onboarding processes and first steps inside the team.

Languages & Frameworks:

Platforms and Databases:

Data and ML Libraries:

Other:

In the context of a data crowdsourcing company, worked in the squad responsible for ensuring the quality of the data submitted by the crowd.

To achieve this goal, worked towards the productization of metrics and models which allowed taking dynamic actions as part of the data flow and building analytical tools for monitoring purposes.

Other tasks performed:

• Improve CI/CD and packaging of artifacts from data scientists and internal tools.
• Development and maintenance of end-to-end tests.
• Helping data scientists in their day-to-day routine.
• Maintenance of micro-services and throughput enhancement.
• Monitoring and alerting of infrastructure.

Languages & Frameworks:

Platforms:

Data libraries:

Databases:

Other:

As part of the e-learning team, I help to solve the emerging challenges through making tasks as:

• Data visualization and exploratory analysis using code notebooks.
• Development of proofs of concept applying state-of-the-art methods.
• Support to design architectures to deploy the chosen solutions.
• Development of machine learning and deep learning models using both classical tools, big data and streaming platforms.

Some of the ML/DL techniques I have applied to projects are:

• Pattern mining and frequent itemsets mining.
• Hidden Markov Models and LSTM Networks.
• Recommender systems using ALS, autoencoders and embeddings.
• Anomaly detection using clustering techniques (kMeans, DBSCAN).

Languages:

Platforms:

Data libraries:

Machine Learning/Deep Learning libraries:

Databases:

Other:

Co-foundated, designed and developed of a collaborative platform for students oriented to the early creation of a professional profile. For this, designed and early developed a microservice-based architecture, by using the Netflix OSS stack over the Spring framework: Zuul, Hystrix, Ribbon, Eureka, Spring Stream...

Technologies & languages:

Maintained and developed a management application broadly used on call centers.

Technologies used:

First contact with a multidisciplinary development team through a collaboration in this startup as an Android developer implementing UX/UI designs.

Technologies used: