Raymond Peck Email & Phone Number

Redwood City, Ca, Us

Palo Alto, California, Us

Progressive Ventures is focused on early stage B2B AI startup investments that have come exclusively through the Y Combinator, StartX (Stanford's venture arm), and SKYDECK accelerators.

Newport Beach, California, Us

Intertru is a method-driven interviewing accuracy platform that uses AI and the HireOS® methodology to ensure that your company hires the strongest person every time.I joined initially as an AI /ML advisor and back end IC, but quickly transitioned into being CTO.

Los Altos, California, Us

The world's largest Industrial Knowledge Graph, capable of finding companies and employees you could or should do business with, in seconds.Was the entire engineering team for the first 10 months, then brought on a small outsourced team.Built all of the back end REST API (Python, FastAPI, SQLModel, Pydantic, SQLAlchemy) and initial deployment infra (AWS, Kubernetes).

Sunnyvale, California, Us

Eta Compute's platform helps vendors train Computer Vision models for edge devices such as security cameras.Introduced static analysis tools, ruff and mypy, into a large Python codebase. Conducted a substantial refactoring effort (36,000+ lines of code) to improve type safety and proactively identify potential issues, resulting in a more resilient and maintainable codebase.Built runtime tooling leveraging Python’s introspection capabilities to strengthen the platform’s extensive configuration system. Implemented runtime type checking to validate complex configurations, significantly reducing error rates and enhancing overall system reliability.

Irvine, California, Us

As the Data Science/Machine Learning subject matter expert, I was the de facto Product Owner, responsible for guiding the engineering and product teams (~60 people) in the creation of Alteryx Machine Learning and the company's underlying Open Source ML libraries (EvalML, Featuretools, and Woodwork). The goal of Alteryx ML is to bring ML to data analysts and domain experts in an easy-to-use yet statistically valid way.All of the DS/ML functionality went through me. I proposed most of the new features, and guided all of the rest. It was my job to ensure that the Data Science of the product was sound, to propose, research, and prototype new functionality, and to develop and present all of the educational content in the product. As part of this, I created and presented courseware for the engineering team and presented to oversubscribed audiences at the Alteryx Inspire conference.I served as the technical contact between the engineering team and customer-facing groups in the company, including the Chief Data Analytics Officer and his team of Data Scientists, and the DS/ML sales engineering overlay team.I also created and maintained a repository of over 2500 benchmark datasets, a metadata database, and an exploratory data science environment. This allowed the team to quickly find appropriate datasets to prototype and evaluate new product functionality.

San Mateo, Ca, Us

Led the 50+ person 3-continent engineering team that implemented all core ML functionality, including automated feature engineering and model tuning (AutoML). I reported directly to the CEO, and engaged ~15 recruiters to expand this team in San Mateo and Warsaw and to bring on a Head of QA and a tech writer. Proposed a new product feature that led to the largest accuracy improvement in v2.0 compared to prior releases.Designed a system to measure and track accuracy and runtime performance to systematically identify the biggest gaps in the company's ML automation and to measure the effectiveness of proposed improvements. The prototype of this system identified a major accuracy regression before release.Reduces release cycles from >6 months down to 2 months by making changes to the engineering team and working on a project to reduce Continuous Integration failures from very high levels.

Palo Alto, California, Us

Lead a small team of junior- and mid-level engineers in the platform work on Quantifind’s new financial crimes investigation products and worked as an individual contributor in Scala. Implemented and supported a networking ETL / Data Engineering subsystem to reliably and interactively connect with a mission-critical data source with more than 1100 unreliable endpoints: adaptive dynamic load balancing, purgatory for temporarily-failing endpoints, and benchmarking and monitoring code for ops alerting and charting. Integrated a new vendor data source for person and organization search into the company’s generic data architecture (Scala ETL / Data Engineering). Implemented an external-facing search API to allow case management software from business partners to integrate with Quantifind’s real-time entity search. • Scala

Developed the specification and implemented validation services for a data ingest REST API for telemedicine data, based on HL7 FHIR. Physiosigns uses this data to build Machine Learning models which predict patient health risks and the actionable factors that affect them. • Python

Mountain View, Ca, Us

Reported directly to the CEO (acted as VP Eng), running the best engineering team I had ever worked with, and made significant code contributions to the code base in Java and Python. I had 15-20 reports, including specialists in Machine Learning algorithm implementation, distributed and massively parallel systems, data science, web front-end, DevOps, and QA. Led a complete redesign and reimplementation of the Machine Learning back-end and released it at 4 pm on the day we had planned over a year before. We then kept up a regular cadence of scheduled releases for the remainder of my leadership: fix releases every 2 weeks and stable feature releases every 6.Drove engineering hiring as well as improvements in much of the back-end architecture and QA.Negotiated and nailed down feature requests from customer-facing data science teams, tracked and prioritized hundreds of feature and bug tickets, ran daily standups, etc.

Mountain View, Ca, Us

Architect and key implementor for the back end of the industry-leading open-source Machine Learning platform H2O (https://www.kdnuggets.com/2018/02/gartner-2018-mq-data-science-machine-learning-changes.html).Designed and implemented all the backend code for H2O's AutoML automatic machine learning facility and its dependencies (Stacked Ensemble models, random hyperparameter search). Designed and implemented simple, usable, and maintainable Java base classes underlying its predictive models and their training (Model, ModelParameters, ModelBuilder). Designed and implemented a framework for the REST API and implemented most of the endpoints which are used by the Python, R, and web clients. This framework makes extensive use of Java annotations and reflection to eliminate boilerplate code while still allowing evolving backend code to present a stable client API. It also presents rich metadata for all API schemas and endpoints, allowing me to auto-generate code bindings for Java and Python clients using a small amount of Python code, and for the front end engineer to write a model builder GUI which automatically adapts to changes in the ML algorithms.

San Mateo, Ca, Us

Designed and implemented the customer REST API (Java / JSON) and SDK (Python), and worked on numerous other features and fixes, including building the landing page for the product. • Java, Python, CoffeeScript (Javascript), HAML (HTML), SASS (CSS) KnockoutJS, Google Closure, PostgreSQL, Hibernate, Scala

Reported directly to the CEO (acted as VP Eng). Lead the engineering team for our live video streaming product, a team of 7 engineers. This was half of the company's product line.Responsible for a worldwide SaaS system of hundreds of servers in dozens of colo facilities running a strict realtime application, and a near-realtime analytics system for customer-facing and internal dashboards and NOC troubleshooting.Radically improved team morale and productivity by creating a culture that empowered the team and greatly improved communication within the team and with the Ops and VoD Engineering groups.Lead successful releases of a complete rewrite of the streaming system, which increased scalability in some cases by hundreds of percent, and several sub-products (Web Services, realtime statistics, Multiscreen delivery, analytics. . .). Drove the establishment of automated testing (load / system / unit) within the group and worked with the QA group to integrate these into Jenkins Continuous Integration. Increased integration with the VoD group by "exporting" the Live realtime stats system to other uses. Drove new projects to extract and analyze network performance and anomalous-event detection from the streaming system for our Network Operations Center, to rewrite the Live archiving system and integrate it into the VoD system, and scale the database.Re-architected our content delivery analytics system to allow BI-type analysis using an RDBMS OLAP cube.

Winner, DEMO Conference DEMOgod award, 2008.This is a proprietary ultra-high-performance live video streaming system serving thousands of streams to hundreds of thousands of concurrent users globally. I built the gateway that allows streaming media ingest into the system from third-party RTMP, RTSP and RTP encoders, designed and implemented much of the data architecture which supports dynamic config of the system and reporting of real-time behavior (big data streaming analytics), and a bunch of other fun stuff. • Java, PHP, MySQL, FreeBSD, JasperReports / Relational OLAP / Mondrian, JavaScript

Louisville, Ky, Us

Transferred domain knowledge and the Jacent source code to a new software team at Yum Brands after they purchased our technology.

Us

Developed and implemented the architecture for a rich voice and web ordering application, including the backend data model and the plugin architecture for connecting to complex third-party ordering systems. System supported multiple customers including Pizza Hut, Quizno's, and Dominos through automated voice ordering, a Swing UI for live customer service reps, consumer web ordering, and an admin-and-reporting (analytics) website. Our hosted call-center solution handled tens of thousands of orders per day with thousands of locations and millions of customers.Wrote large parts of the application.Worked with the ops team to deploy and support the application at hundreds of remote sites with many customers.Managed an offshore development team of a team lead plus 6 other engineers. • Java, Spring, Hibernate, MySQL, Swing, Linux, JBoss, Tomcat, VXML, NT Embedded

Us

Participated in first- and second-round financing.

Us

Architect for inSite, a hosted customizable personalized web portal / eCommerce site developed for our large corporate customers (HP, Earthlink, Sprint et. al.). Portal included our own consumer eCommerce product, B2B integration, third-party data feeds, etc.As technical product manager, provided technical contact between these customers and our engineering team.Lead several engineers in developing the product. • Java, MySQL, HTML, XML, SOAP

Technical contact between the company and our large corporate customers (HP, Sprint, Earthlink. . .). Drove technical requirements, and lead the effort to implement those requirements, including developing project plans and managing a team of junior engineers.Specified and implemented application features, our proprietary language interpreter (WebC), scalability and portability issues, interfaces to third-party systems (provisioning, system monitoring, online payment processors, etc.) Integrated our product with a number of others to create a turnkey eCommerce platform, which HP offered to large portal customers including BellSouth, Telefonica and Interland. In charge of all technical aspects and all non-financial communications with HP and our mutual customers.Wrote an XML-based high-performance hosted website development tool and infrastructure (SiteBuilder) for customer Earthlink. Extended WebC language to natively support XML and bean-graph <-> XML mapping. • C, WebC, xBase, HTML, XML

Worked in the platform team to develop a CORBA-compatible movable-object infrastructure in Java. Objects running on a set of loosely-coupled servers supported remote method call and the ability to move their state between Java virtual machines over the Internet. Implemented a distributed naming service and registration, application-startup server discovery, and a pluggable, extensible remote-debugging framework for client objects. The applications group used this platform to create applications for distributing a hotel reservation system for Hilton, and a sample application for billing/invoicing. • Java

After our IPO I joined the sales organization for a year to get more experience communicating with customers from a business perspective.Presented complex configuration management solution (ClearCase) to potential customers, and supported the product during evaluation cycles. Managed expectations and requirements of potential customers in a pre-sales setting, communicating these back to the remote engineering team and management.Exceeded sales quotas each quarter. :-)

Founder. Helped company grow from 2 to over 200 and have a very successful IPO (~$750M valuation) in 5 years. Pure is now part of IBM.Key developer of Purify, Jolt Award winner (Software Development Magazine / Dr. Dobbs).Key developer of Quantify, Java Developer's Journal award winner, Jolt Award winner (Software Development Magazine / Dr. Dobbs).Purify was the first product ever to parse and rewrite object code. It used this revolutionary technology to pinpoint errors in compiled code and inside third-party libraries. I worked on all areas of the product except the GUI.Quantify was the first cycle-accurate code profiler, the first to accurately track the runtime call graph, and the first to have a simple, powerful GUI. It used object-code instrumentation to count precisely the number of machine instructions that a program executed, in real time. Worked on code analysis and generation, CPU pipeline simulation, ran the usability studies for the GUI and drove UX improvements. • C, SPARC assembly language, SunOS executable file formats

San Jose, Ca, Us

Member of the 3 1/2-person team that developed the MIPS-architecture CPU that powers the Sony PlayStation, PlayStation 2 and PSP (over 200 million sold). I designed and implemented the complete instruction decoder (which generated control signals for the rest of the CPU) and the instruction-fetch logic, was in charge of QA for the ALU, and built behavioral models to act as a test harness for other parts of the CPU.Chip taped out successfully the first time with 0 bugs in the CPU.Author of one and co-author of another IEEE paper about this project. See below for more details.Developed custom CAD software which was used by the 6+-person chip-design team to support logic design. See IEEE paper for more details.Developed a low-level custom graphics library for an asynchronous pipelined graphics accelerator, for use in hardware X-terminals using our follow-on chip.

Ms, Electrical Engineering (Computer Architecture Specialty) And Most Of An Mscs

Bs, Electrical Engineering