cv
General Information
Full Name | Daniel Pereira da Costa |
Languages | English, French, Spanish, Portuguese |
Education
-
2024 Master of Applied Data Science
University of Southern California -
2019 Master of Engineering - Double Degree with PUC-Rio
École Centrale Paris -
2019 Bachelor in Control and Automation Engineering, Minor in Mathematics
Pontifical Catholic University of Rio de Janeiro (PUC-Rio)
Experience
-
2023-Today Data Analyst
Intel Corporation - Develop an all-encompassing PowerBI dashboard for proactively identifying performance deviations in 50+ AI workloads running on Intel's Ponte Vecchio GPU across single-card, multi-card, and multi-node configurations.
- Design a PoC using GPT-4 and Llama-Index to ingest and associate unstructured customer engagement updates from disparate sources (e.g., PowerPoint, JIRA, Excel) into a single PowerBI based reporting for several high priority, high visibility engagements
-
2022 Data Engineer
Akad Seguros - Orchestrated design of 600GB Data Lake on AWS through S3, Glue, Lambda, and Step Functions; published to AWS Blog Brazil.
- Ingested over 100GB of data from 8 different on-premises and cloud-based servers into Data Lake via Data Migration Service and ETL jobs.
- Defined and deployed Data Warehouse (50GB) components and architecture on AWS applying tools such as S3 for storage, Spark for processing, and Athena for analytics.
-
2020 - 2021 Data Scientist
Cyberlabs (Google for Startups 2020) - Engineered Machine Learning model with 71% f1-score to reduce churn from 6 million user cybersecurity app; deployed model in AWS utilizing EMR Clusters with Spark for data preprocessing, EC2 instance for ML inference, and Data Pipeline for ETL.
- Guided a team of 4 in creating a Churn Predictive Model (LSTM network) with 72% precision and 83% recall for telecom company; leveraged statistical analytics techniques to analyze customer churn patterns against 10 key metrics.
- Spearheaded a team of 3 in constructing a system aiding 20+ companies in controlling COVID-19 spread for 3,000 employees; implemented event-driven architecture on AWS employing Infrastructure as Code and Lambda, SQS, SNS, API Gateway, and RDS.
- Collaborated with 2 members to construct an XGBoost model with 77% f1-score to maximize customers' revenue.
-
2019 - 2019 Data Scientist
Cyberlabs (Google for Startups 2020) - Built Node.js back-end service generating up to 10 insights er minute from Computer Vision Model; increased 100 establishments' decision-making productivity by 20 hours/month.
- Integrated and deployed microservice with GraphQL and Docker.
-
2018 Embedded Software Engineer
Softbank Robotics Europe - Presented reports to firmware team to validate sensor's effectiveness against 2 KPIs.
- Created firmware module in C/C++ embedded in a microcontroller for a robot's inductive sensor (LDC1312/4).