Jaime Espinosa

Cortex (AI/ML)

Compute for AI/ML: Serverless platform and development tools for productizing AI/ML ● Led the product re-design of the platform, end-to-end ● Owner of the public PaaS offering; used to incubate products for the enterprise SaaS · Expanded target market ~50% by adding new segments and improved product-market fit · Defined product vision and roadmap by driving consensus across CTO, VP's, PM's & Eng. · Kicked off 8+ mo. of development by delivering product requirements docs; MVP to v3 ● Assessed competitors, partnerships, and leverage opportunities by diving deep in AI/ML

AI + Data Platforms for Sales: Recommendation engine. Reinforcement Learning. ● Delivered the first version of a suite of AI for sales enablement: inc. by 30%, $4B+ in year 1 · Architected the UI and machine learning platforms together; for a UI and AI feedback loop ● • Drove 5 sales-and-marketing products that continuously optimize revenue with ML · Designed: the user experience, and defined architecture, data/feature selection, & models · Developed revenue optimization: GA on content, salespeople as mutex, & fitness on revenue ● Owned user adoption and impact on total sales revenue for each of the products in the suite · Delivered “Business Value Programs” - Presentation, Content mgr, Calculator, Demo, & CRM

Compute Platform for IOT: Serverless for internet-of-things. Startup ● Built proof-of-concept and secured funding for a product that brings “serverless” to IOT · Wrote a library and core service for firmware developers to use cloud for processing

Compute Platforms for Cloud: Serverless and App Platform as a Service - distributed system ● Lead the first-to-market Serverless product - Azure WebJobs (renamed Azure Functions) · Managed the end-to-end vision and implemented it through influence and partnerships ˙ Led the team of PM’s, engineers, designers, and writers to build a cohesive product ˙ Created a product plan for Azure’s serverless, using ML to analyze customer’s patterns ˙ Managed partnerships: Visual Studio, App Service, Scheduler, ASP.NET, Azure SDK, UX ● Developed Serverless POC, business model, presented to 3 leadership layers of 6 groups ˙ Built a community of users with growth of 20% month-to-month, b/c customer-obsessed ˙ Designed pricing, monetization, profitability, segmentation, and competitive analysis ˙ Drove the marketing plan, branding, public relations, communication strategy ● Drove infrastructural architecture of Azure App Service platform for reliability & scalability · Performed an architectural audit, drove major changes, established processes · Delivered a scalable traffic routing security layer. Developed automated capacity mgr. · Released a new core system to increase parallelism, scaling, and availability; adding a “9” · Established release processes to ensure customers were never affected 

ML + Data + Development Platforms for products: Analysis & BI modeling and Data Engr. ● Built and lead the internal Analytics team to enable data-driven business decisions · Presented in weekly to 4 tiers of management to identify opportunities, risks. Helped GA · Influenced the way Azure did market targeting, system health, growth hacking · Delivered a big-data warehouse and reporting for business decisions – POC for larger org · Drove the implementation of analytics tools, and ML tools – a “product science” service ● Drove Remote Debugging and Remote Profiler features and services · Integrated groups: Microsoft Security, Visual Studio, Azure SDK, App Service. 30% of users ● Drove “Testing in Production” – First Azure App Service testing-as-a-service product · Designed and released a routing and testing platform, adopted by +75% of users · Enabled users to optimize applications by providing a platform for A/B testing, analysis, ML

● Scrum master for Azure Web App business group - driving all projects’ priority · Drove 20+ product cycles, and feature triage for the whole App Service group (60 ppl) ● Managed Documentation and Marketing for groups: API apps, Web apps, Mobile apps, Serverless · Managed teams of 20 writers and 30 PM’s to consolidate and document the platform · Worked with the marketing to coordinate and target messaging and generate demand

Compute Platforms: Serverless distributed system research – compute as a commodity ● Research on Cloud Computing, focused on distributed computation · Parallelization of Heterogeneous Computing Environments Through Simulation (of CPU's) · Researched methods for an MPI multicore accelerator by modifying a hypervisor · Developed virtual I/O devices for our hypervisor: Palacios Virtualization Platform ● Research on Internet of Things / Wireless Sensor Networks · Collaborated in various WSN papers to aid IOT development, and distributed computing · Designed and built a “Power Harvesting Module”, able to harvest power from any source · Developed a big-data system with for acoustic fingerprints for a mobile app back end ● Adviser: Peter Dinda, Ph.D. – Head of Systems Engineering at Northwestern University

Leading teams: physicists, hardware engineers, statisticians, material scientists. Qual. & Rel. ● Launched processor products, driving them through the qualifying and manuf. stages · Developed strategies and managed tactics to ensure schedules were met – per stage · Worked with customers to understand their needs, find solutions, and keep schedule ● Owned CPU products and brought to market 2 product lines ˙ Itanium: Intel's first 64-bit, natively parallelized CPU: Tukwila (93xx), Poulson (95xx) ˙ Atom: Silverthorne: Intel's first IOT CPU to market. (Z5xx), and (N2xx, 2xx, 3xx), ˙ Atom: Pineview: Intel's first system-on-chip to market. (N4/5xx, D4/5xx)

AI Platform for hardware: Data platform for CPU design/manuf. and ML tools. Statistics. ● Created processes and modeling methods for CPU and Chipset groups; design to manuf. · Transcribed from math to code: physical, parametric, statistical, and behavioral models · Influenced various groups by providing data and analysis of the impact of new methods ● Developed an AI-based optimization tool for use on product life-cycle simulators · Modeled process manufacturing, product features, and usage to build full product models · Implemented by combining particle swarm optimization, fuzzy logic, & genetic algorithms · Used by CPU and chipset product groups to improve yield by ~7% average, used by design ● Evangelized software products and business methods to promote adoption · Presented white-papers, demos, and classes at conferences and Intel university · Published a number of internal Intel Publications (analogous to academic white-papers)

Developer Platform: Scientific tools, data collection ● Managed, architected, implemented, and evangelized the use of our scientific modeling · Managed clients and partners to define the requirements, UI tests, formulae, and reviews · Researched and applied new technologies to develop proof-of-concept, and into platforms ˙ Developed API and infrastructure to enable developers to expand model libraries ˙ Built a development platform for coding, distributing, and maintaining a suite of 7 tools

Consultant for startups ● Implemented control system for a robotic arm – medical researcher at Indiana U. (‘05) · Implemented a combination of genetic and particle swarm algorithms for motor control ● Designed and built a prototype for a 3D triangulation of IR source - for solar-panel (‘05) · Built a sensor array to triangulate in real-time an infrared source in 3D space ● Designed & built a prototype for a heads-up-display for Dodge Vipers –racing team (‘04) · Led engineer and product manager to team of 5; responsible for design, schedule, budget · Developed kernel modifications for real-time device communication with the car computer Personal Projects ● Implemented an FPGA within an FPGA: for reconfigurability and experiment with ML (’06) ● Built proof of concept of self-configurable circuits as method for fault-tolerance (’07) ● Making a smart house through unsupervised learning (’14-present) ● Pet Project: Use AI + gamification to find lost dogs (‘18-present)

Data Platform for Quality: data acquisition, automated analysis, and report generator ● Built the quality lab from the ground up with a tight budget. Trained staff on the process ● Implemented a sensor network and embedded system ‘broker’/hub to post to a repository · Automated data acquisition, analysis, and reporting for Six-Sigma and QS9000 · Enabled an increase of parts sales by 2x