Back-end Go Developer
2019 - 2020Calm.com (via Toptal)- Separated subscriptions from the monolithic application into Go microservice with a CQRS/Event Sourcing architecture.
- Designed subscription data models for production and pipeline to the analytics team.
- Developed a back-end for promotional pricing with A/B testing.
- Improved test resolution/ease-of-use, contributed new patterns for more agile and less brittle integration tests, reduced docker image sizes 10x, feature enhancements to in-house DevOps tools.
Technologies: Amazon Web Services (AWS), CircleCI, Jenkins, Terraform, Kubernetes, Docker, Datadog, LaunchDarkly, Redis, PostgreSQL, Apache Kafka, AWS Push Notification Service (AWS SNS), Amazon S3 (AWS S3), AWS, Amazon SQS, Node.js, JavaScript, GoDirector of IT (2014-2019) | Lead Engineer (2009-2014)
2009 - 2019Columbia University- Developed a comprehensive services management and billing system with multiple web front ends that securely made available 30+ million records per year, resulting in a frictionless UX (React, SSR, server streaming, Vue, gRPC, Go and PostgreSQL).
- Productized research group's method of identifying protein-protein interactions with an improved algorithm (1,000x faster). This speed-up enabled them to stay within budget and explore 100x more genomes (built with Rust, Python, C/C++, SQLite and ASM).
- Designed and coordinated the implementation of the architecture for secure sharing and analysis of patient data. Was responsible for on-going auditing and oversight of security compliance (HIPAA and internal controls).
- Cleaned and restructured a 40 TB genomic dataset for a quality and high-performance distributed analysis in the cloud (AWS) along with a custom storage engine.
- Created an original dual-space algorithm (41 years reduced to 18 hours) to identify complex genetic disorders (built with ASM, GPU).
- Built a container system to sandbox users and fair-share limited resources (built with C).
- Developed a compute swarm that robustly handles failures, dynamically redistributes computation, and scales efficiently—scaled to 6,000 instances (built with C, Go, and Python).
- Wrote a Linux overlay file system with real-time adaptive, user-based QoS limits for a 5,000 node compute cluster to prevent runaway usage. A user can only slow themselves down (built with C).
Technologies: Amazon Web Services (AWS), MySQL, PostgreSQL, Google Compute Engine (GCE), AWS, Kernel Programming, ASM, Graphics Processing Unit (GPU), MATLAB, R, Protobuf, gRPC, Python, C, C++, Rust, Go, JavaScriptSenior Vice President
2003 - 2004Lehman Brothers- Led profitable quantitative equity trading team within a mixed disciplinary group.
- Removed an obstacle to a strategy testing framework by adding time-series extensions to MySQL and improved strategy development time three times.
- Wrote an Excel extension to support the real-time bidirectional communication and live-feed of market data.
Technologies: Perl, MATLAB, R, C, C++Vice President
2000 - 2002Morgan Stanley- Proposed, modeled, and implemented improvements to VWAP execution strategy yielding a 3% price improvement without notably increasing volatility.
- Developed (full-stack) a monitoring system to track all of the firm’s agency equity trades flowing through our execution strategies internationally.
- Built a robust C++ live trading engine with an embedded JavaScript VM for Nimble and safe incremental execution strategy development.
- Increased customer utilization by 10% by tuning and extending the limit-order model.
- Increased returns by 5% on proprietary reversion strategy by better modeling and anticipating risk.
Technologies: JavaScript, Python, Perl, MOSEK, SAS, MATLAB, R, C++Project Manager
1999 - 2000Investment Technology Group- Refactored and restructured a broken portfolio optimization product to reinvigorate sales.
- Increased the optimizer value by creating and implementing a mathematical model for tax handling of wash-sales.
- Heavy interaction with clients to provide unique extensions for their needs.
Technologies: CPLEX, MATLAB, C, C++Founder
1997 - 1999Pencom New Technologies- Designed and implemented back-end systems and RPC to the front end to manage distributed telephony switching equipment for a client.
- Implemented a two-tier data entry system for trading floor booking for a client.
- Grew the consulting startup to 35 full-time employees.
Technologies: C, C++Senior Software Analyst
1996 - 1997Aspen Technology, Inc.- Sped-up new process modeling by developing chemical engineering specific scripting language.
- Reduced time spent on DevOps by 25% teamwide by creating a multiplatform, incremental build system.
- Built a CORBA RPC back-end for chemical process modeling and optimization software.
Technologies: CORBA, Python, Perl, CLead Unix Administrator
1994 - 1996University of Denver- Developed a frictionless portal to bill 10,000 students for printing services.
- Built a DNS front-end/database system for managing endpoint networking for the whole campus.
- Managed systems administration team for all central Unix systems.
Technologies: Print, Printer Drivers, RPC, Perl, C, Unix