Resume

Joined a team building agentic commerce infrastructure during organizational uncertainty (minimal docs, no onboarding, significant inherited tech debt). Operated with high autonomy, identifying systemic gaps and driving solutions without direction:

• Designed and implemented a Smithy schema registry to centralize data models across Scala, TypeScript, and Go services, eliminating model drift and parallel definitions.
• Resolved chronic dependency conflicts across core services and jobs by authoring shared libraries and build plugins, removing a recurring source of production incidents.
• Fixed connection pool mismanagement in Beam tasks (orchestrated with Airflow) that had silently dropped millions of records/day for years; defined an observability approach for the full ingestion pipeline.
• Introduced E2E testing for critical APIs; stabilized multiple legacy projects from non-compilable/broken states to deployable, unblocking other engineers.
• Submitted improvement proposals the team adopted: shared schema registry, E2E tests for the full pipeline, continuous delivery, Beam resource management.

Worked on the notification platform delivering real-time messages to SiriusXM apps and devices, a multi-region system handling around 20,000 requests per second. Owned several backend services, serverless components, and the AWS infrastructure around them.

• Reduced the team’s AWS cloud costs by 80% while preserving throughput.
• Engineered the foundation for a zero-downtime migration to a new AWS account, so the team could continue the transition seamlessly.
• Contributed to the design and implementation of the notification system: event ingestion pipeline, shared libraries, autoscaling policies, and observability.
• Designed the initial framework to run end-to-end and load tests against the notifications pipeline.
• Restored critical legacy Java systems unmaintained for 5+ years, from non-compilable code to a deployable state, and migrated them from Spring Boot 2 to 3.
• Implemented the CI/CD pipelines (deploy/hotfix/rollback) coordinated with the E2E suites, built a reusable Scala template for AWS Lambdas, and designed an internal Slack bot the QA team used to run common tasks against our services.

Joined a 5-person R&D team building enterprise blockchain products. The team shifted focus to self-sovereign identity and grew past 40 people on the technical foundations I designed:

• Designed and led most of the foundations for Atala PRISM (now Hyperledger Identus, https://identus.io), a self-sovereign identity (SSI) ecosystem built around Decentralized Identifiers (DIDs) and Verifiable Credentials.
• Designed the initial second layer protocol for storing decentralized identifiers and digital credentials on top of Bitcoin, and the initial architecture for the Credentials Verification project.
• Designed and implemented the first PRISM SDK, cross-platform so web, mobile and backend clients shared a single implementation.
• Built the API documentation website so that every code snippet was compiled as part of the build, which meant the examples could never drift from the SDK.
• Coordinated external vendors delivering our initial mobile and web apps, and wrote technical design proposals for prospective customers.
• Presented PRISM at public events, on-site in Mexico (Viva La Crypto, Talent Land) and online: https://www.youtube.com/watch?v=8AbgzvOeb6w

Worked on ARTIK Cloud (now SmartThings Enterprise), an IoT platform to connect devices and applications.

• Implemented Amazon S3 resiliency across regions for the image storage service, so a regional S3 outage no longer takes the service down.
• Implemented TCP and TLS support for the LwM2M protocol used for IoT device management, and backported fixes from our forks into upstream Eclipse Leshan and Californium (leshan, californium).
• Contributed to a multiple-active-regions project using multi-master asynchronous database replication across regions worldwide, using Conflict-Free Replicated Data Types (CRDTs).
• Contributed to the Device Management Approved List, which lets an organization whitelist the devices allowed to connect to the cloud.

• Helped in the architectural design for the core of a big data pipeline processor.
• Designed a domain specific language for type safety while writing Hive queries and Spark transformations.
• Created a code generator for our domain specific language (based on Hive tables).
• Redesigned how to integrate the data pipeline processors.
• Improved team coding practices to get better benefits while using Scala.

Fintech lender offering online micro-loans. I was part of the team running a critical migration from a monolith to microservices, which also moved the platform from MongoDB to MySQL.

• Participated in the design and implementation of that migration, which scaled the platform from hundreds to thousands of concurrent clients (https://kueski.com/).
• Designed and implemented an internal REST API to screen applicants against criminal records and politically-exposed-persons (PEP) lists, storing enough auditability to reproduce and understand each query result.
• Designed and implemented an async web service to rate-limit the volume of traffic sent to an external provider to avoid overloading it.
• Solved two bottleneck issues stopping the loan-request process (caused by external providers).
• Helped to architect a secure system modelled on PCI compliance requirements.
• Designed an automatic database updater inside a RESTful service with short downtime without DB transactions while keeping consistency.
• Supported external Android devs (domain knowledge, reviews, standups).

• Maintained a module generating metrics about RAC (Real Application Clusters) in the Oracle database.
• Helped design a module for compressing information used to reconfigure the RAC cluster.
• Helped implement compression of log messages for the Oracle database.

• Juicio En Línea Sinaloa: a government web app letting citizens sue government dependencies, with the whole judgment process handled online.
• Created applets using biometric devices to capture digital fingerprints.
• Boosted a government application’s performance by fixing concurrency issues; fixed several web vulnerabilities and redesigned some legacy; integrated version control into the dev environment.