Carmine Michael DiMascio

Led a cross-functional team to architect and implement a modern, scalable solution that powers all Tripadvisor restaurant search capabilities.

The solution uses the CQRS pattern. Data is published to and sourced from multiple event streams, then aggregated and optimized for search. The service exposes both RESTful and GraphQL interfaces. It's backed by Kafka, Postgres, Memcached, and Elasticsearch.

Led a team to implement a custom service mesh based on go-control-plane, Envoy xDS, and GRPC. The mesh empowers teams the flexibility leverage 'off the shelf' technologies while enabling communication with Tripadvisor's proprietary service stack and monolith.

Tasked with the mission of closing the "30 million word gap", I  lead a team of 6 engineers and work closely with IBM Research to develop and deliver an innovative new API platform, Watson Tutor for Early Childhood Learning. This work was featured on the Discovery Channel episode, "This is A.I.". The platform is a set of APIs and services that enable developers to create personalized, adaptive learning experiences for early childhood learners.  I also lead the development of two exemplar applications built atop the platform; "Ernie", a smart toy prototype and Word Power Ups, a cognitive vocabulary iOS application focused on classroom learning.

Watson Tutor is implemented as a modern microservice architecture on Kubernetes. Microservices are implemented on stacks including Node.js, Spring  5  Kotlin, and Go. It uses services such as RabbitMQ, MongoDB, Postgres, NGINX, Watson, and more. 

Word Power Ups is a native iOS application implemented in Swift.

"Ernie" is a smart toy, prototyped on Raspberry Pi  with Javascript, C++, and Watson.

I am the lead engineer and core contributor for all three projects.

Led a team of 15 engineers to develop the  IBM Watson Advertising platform. Watson Advertising is a cognitive advertising platform that leverages Watson AI to deliver personalized digital advertisements in the form of a smart conversational agent or chatbot. I led the effort to grow Watson Ads from concept to multi-million dollar product within a short time span.  The product has earned industry recognition. It was a  finalist for Best Ad Tech Tool,  MMA Technology and Innovation award, and MediaPost OMMA.

Watson Ads is built on AWS and serves ~50 million requests per week. It is implemented as a modern microservice architecture with services written in Node.js and Java.  It leverages services including Cloudant, Watson, ElasticSearch, and more.

Nominated by Steve Abrams, IBM Distinguished Engineer and Rob High, IBM Fellow to lead the 'technical due diligence' effort during IBM's  review of Alchemy API. My role was to determine whether the Alchemy API technology was viable and in alignment with our strategic imperatives and recommend whether or not to acquire. Ultimately, my efforts played in a significant role in IBM's acquisition of Alchemy API

I was also nominated co-lead of IBM's 'preliminary technical due diligence' effort during IBM's review and eventual aquisition of Blekko.

Led the IBM/Kellogg engagement, a collaboration to  deliver the Bear Naked Custom Granola web application powered by Chef Watson. Bear Naked Custom Granola enables users to create novel, tasty granola mixtures with Watson that pair well at  both the chemical and molecular level. The app appeared on the Today show it December 2017.

Led the development of  the Chef Watson API. An API which powered Kellogg and Watson Twist. It is implemented with Java and Node.js. It is hosted on Softlayer and delivered as a CloudFoundry service. The API leverages services  including Cloudant, IBM DB2, Watson, and more

Core contributor IBM Chef Watson, a cognitive AI system trained on recipes from Bon Appetit and capable of generating unique and novel recipes from scratch. Recipe ingredients, proportions, and steps are all generated by the system resulting in never before seen recipes that work. Chef Watson placed eighth in the 2014  Fast Company "Innovation by Design" award. First place went to Slack. Chef Watson also generated ~1 billion media impressions and over 300 articles including Tech Crunch and WSJ.

IBM Hackathon Winner for the development of the Watson News application. It extracts key entities and concepts from unstructured news articles and provides summaries, timelines, and links to related content.

Led a small team of 5 engineers to create a data analytics service and data lake using Apache Spark, Scala, and Python.

Led a small team of 5 engineers to develop and deliver  Watson Twist, a cognitive cocktails application for iOS and Android. The application was built using React Native.

Collaborated with Marshall Schor, chair of the Apache UIMA project and members of IBM Research to transform a UIMA based medical information annotator into an easy to consume, IBM Bluemix RESTful service. I transformed the annotator into a cloud service and implemented both its CloudFoundry service broker, and frontend demonstration application.

Technologies included Node.js, AngularJs, CloudFoundry, and Apache UIMA.

Core contributor to the Rational  ClearCase, Rational Team Concert, and Rational Software Architect products. I worked on both the clients and server. Each client was implemented in Java as an Eclipse plugin/application. The server was a JEE application that included both C and C++ components.

I spent ~2 months in Stockholm, Sweden working with Ericsson to improve a mulit-million dollar, deteriorating relationship. I worked with key Ericsson stakeholders and developers to improve productivity, identify key problem areas, and deliver features and fixes on-site. I met with IBM and Ericsson IBM senior management to ensure KPIs were met. Finally, I led two education sessions, each of which was attended by ~200 Ericsson engineers.

Masters of Arts in Computer Science

Magna Cum Laude

Minor in Mathematics and Western Civilization