Experience
2025 — Now
Redwood City, California, United States
• Building LLM-powered systems and agent-based infrastructure to advance observability intelligence.
• Designing scalable AI features for production using Amazon ECS, ML pipelines, and internal platform tools.
• Transitioned to AI Innovation in April 2025 after contributing to early internal AI initiatives.
2023 — 2025
Medellín, Antioquia, Colombia
• Achieved a 98% reduction in on-call noise, cutting monthly pages from over 270 to just 3–4 actionable alerts by overhauling processes and streamlining workflows, significantly improving team focus and operational efficiency.
• Attained 100% SLA compliance for P1-P3 closed bugs in Q2 2024, increasing from 50% in Q3 2023, through process optimization, workflow automation, and accountability initiatives.
• Delivered 10+ high-priority initiatives over three consecutive quarters while maintaining team performance during leadership transitions, mentoring a new Engineering Manager to ensure seamless continuity.
• Drove 90% adoption of Open Integration Manager (OIM) across 100+ active integrations, launching OIM-enabled solutions, delivering automation integrations (AAP and EDA Ansible), and improving debugging with audit logs and human-readable timestamps.
• Transformed the software development lifecycle (SDLC) by introducing standardized deployment processes, building robust preprocessing functions, replacing legacy systems with a new integration API, and empowering technical writers to independently update in-app documentation through a CMS.
• Led hiring efforts, conducted performance reviews, and supported career development planning across the team, while mentoring a new Engineering Manager during a leadership transition.
Note: Transitioned from dual Tech & Team Lead to Tech Lead in Aug 2024 following a reorg and to support green card relocation, with Velocity Global S.A.S serving as employer of record for legal and immigration compliance.
2020 — 2023
2020 — 2023
Mountain View, CA → Medellín, Antioquia, Colombia
• Designed, scoped and developed a near-real time ETL Pipeline for BigPandas platform resources using Fargate services, SQS, S3 and Redshift that handled ~2 million records per day in 7 minutes on average that allowed customers to generate dashboards and reports.
• Developed preprocessing services, facilitating seamless data normalization for 80% of incoming data through concise and reausable functions, enhancing BigPandas Open Integration Manager functionality.
• Provisioned and deployed resources across multiple regions in AWS through the use of Terraform to enable high availability and recovery disaster for all of the integrations and services.
• Improved the functionality of multiple integrations such as PagerDuty, Grafana, Datadog, Prometheus, New Relic, Dynatrace, and Cloudwatch by adding improved logging, better normalization of alerts, and adding terraform to handle the serverless infrastructure.
• Designed and developed an integration with Splunk Enterprise and Cloud that allowed users to bring their alerts information into the platform by using AWS Lambda, ES6, Terraform, and python for the frontend.
• Extended the functionality of an internal tool called kungfu that streamlines the safe packaging and deployment of integrations to the cloud by providing infrastructure plans to prevent mistakes by leveraging Terraform, ES6, and AWS.
Note: Employed via Velocity Global S.A.S. from August 2020 to November 2023 for legal and immigration compliance during relocation to Colombia.
2019 — 2020
2019 — 2020
San Mateo, California, United States
• Distilled the vision of the company by collaborating with a cross-functional team on the go-to-market strategy, the company's personas and the product roadmap that helped the company pivot and get traction with customers.
• Optimized the release cycle by implementing product processes, creating a product backlog and doing sprints that allowed us to measure engineering velocity for the first time.
2017 — 2020
2017 — 2020
San Mateo, CA
• Provided benchmarks for users to compare Ople's model by developing a library and micro-service that automatically builds multiple models such as decision trees, random forests, linear regression, logistic regression, and gradient boosting with Scikit-learn.
• Improved the users' understanding of their model by providing them with the features that the model considers the most important from a dataset by implementing the mean decrease in accuracy technique.
• Increased the model's explainability by creating automatic reports for classification and regression problems through metrics such as the Confusion Matrix, Precision, Recall, False Positive Rate, ROC Curves, Precision-Recall Curves, F1 score, MSE, MAE, and RMSE using Scikit-learn, Pandas, and Numpy.
• Implemented the data signature library that is used to predict the minimum model training time for optimal results and the memory needed for model building jobs.
• Increased the models' performance by ~3% by creating a library that creates data pipelines that apply feature engineering techniques such as imputation, frequency encoding, one-hot encoding, target encoding, and data scaling to datasets before training the model.
• Designed and developed integrations between Ople and Tableau by leveraging the Tableau APIs, Ople's API and python microservices that allowed customers to perform descriptive, predictive and prescriptive analytics.
Education
Indiana Institute of Technology
Master of Business Administration - MBA
Baskin Engineering at UCSC