Author Image

I'm Alex Tonetti

Data Scientist, Developer & Entrepreneur

Hello! I’m Alex Tonetti, an experienced consultant with deep expertise in Artificial Intelligence (AI). I Develop data-driven solutions for clients that solve complex problems with cutting edge Natural Language Processing (NLP), Machine Learning (ML), & analytical technologies. I am highly skilled in Python, Java, Javascript, CSS/HTML, etc. Additionally, I am a demonstrated professional with a Masters of Science in Computer Science focusing in ML from Georgia Institute of Technology and Bachelors of Science in Computer Engineering from the University of Maryland, College Park.

Technical Development Skills

Pandas & NumPy

My Experience

  • Graduate

  • 2017 - 2020

    Georgia Institute of Technology

    Masters of Science in Computer Science
    • 3.9 GPA
    • Specialization in Machine Learning
  • 2020 - Present

    Deloitte Consulting LLP

    SensingBridge Solution Architect
    • Served as SensingBridge’s Solution Architect, Data Science (DS) Lead, Deployment Specialist, and briefly the Development Lead
    • Successfully managed a blended team of approximately 12 on and off-shore resources, and external contractors in different time zones
    • Built, managed, and administered the product roadmap across the entire application—DevOps, Front-end, Back-end, API, and DS layers. Evaluated SOW's for client engagements, agreements for data sources, contractors, and internal partnerships
    • Assessed the technical feasibility and created action plans for product initiatives, client requests, and new product use-cases as the Solution Architect
    • Evaluated team members during monthly checkpoint meetings, identifying strengths and improvement areas ensuring timely practitioner feedback
    • Deployed, managed, and monitored over 15 client engagements. Delivered a manual deployment of SensingBridge for a sensitive client requiring the configurations of AWS web services—EC2, RDS, ElasticSearch, S3, Elasticache-Redis, etc.—and the manual installation of backing services— JanusGraph, ElasticSearch, Cassandra
    • Supported and organized sprint ceremonies and related activities including ticket management and creation, sprint review, backlog grooming, story estimation, and pull request review and merges
    • Worked with management to define staffing needs, gave technical interviews, and crafted a code test
    • As the DS team lead, the team explored different techniques for text summarization (abstractive vs. extractive) while improving speed, refined the scenario monitoring strategy, added geoparsing for location extraction, and integrated LexisNexis News
    • Developed technical diagrams illustrating application architecture, data-flow, and component diagrams for non-technical audiences and sales pursuits
    • Mentored and coached the team helping practitioners grow and utilize their strengths. Collaborated with other assets to share lessons learned, best practices, etc.
  • Jun 2019 - 2020

    Deloitte Consulting LLP

    SensingBridge Data Science Team Lead
    • Began building strategic initiatives including componentization of the pipeline for dependency injection, parallelizing data source loading, and distributing the pipeline using Dask
    • Further developed the initialization scan to ensure all DB fields are present in the target DB. Added client customization for ontology root tagging, querying, and query methods
    • Managed and contributed to a dynamic team that collectively achieved:
      1. A custom patent Word2Vec and FastText model for related term generation with a profanity filter, client-defined related term functionality and alternate organization names from CrunchBase
      2. New tagging and scoring methods with increase configurability, letting clients dynamically set tagging and scoring method and scoring impact weights
      3. Exploration of a new method of fuzzy deduplication (locality-sensitive hashing) and combined near-duplicates into a single multi-sources document
      4. Tactically started collecting data at scale, storing all derived metadata from the DS pipeline including tracking discovered entities from named-entity recognition service for the integration of trends
      5. A method for scenario monitoring using parts-of-speech dependency parsing to detect the polarity of ontology elements, utilized sentiment analysis and supervised learning (Support Vector Machine) techniques
    • Improved data querying by adding related terms from Word2Vec and term cleansing
    • Enhanced client engagement tools for efficiency and changing business requirements
    • Collaborated on a cloud-agnostic strategy for client deployments. Deployed, managed, and monitored 6 client deployments
    • Created architecture and data-flow diagrams. Aided in the creation of a DS pipeline overview deck for client teams
    • Mentored and coached the DS team and firm initiative members. Collaborated with other assets to share lessons learned, best practices, etc.
    • Worked on a $100 million proposal pursuit and recruited both for SensingBridge and Deloitte
  • Jan 2019 - Jun 2019

    Deloitte Consulting LLP

    SensingBridge Data Scientist
    • Refactored the code-base using 12-Factor Application and object-oriented programing principles resulting in improved developer productivity, code reusability, readability, and a more fault-tolerant production codebase
    • Improved pipeline space and time complexity increasing performance by over 500% by establishing parallelism and vectorization of functions, and removing unnecessary objects in memory
    • Added new methods for synonym generation, data cleansing, keyword tagging, text summarization, credibility and document scoring, document deduplication, foreign language translation, relevancy scoring, document classification, direct database connection with SQLAlchemy, etc.
    • Built and trained the team on Git and development best practices, created a Jupyter notebook for rapid prototyping and documented the entire application within the code and on Confluence
    • Integrated 88 and evaluated 196 unique data services. Expanded the number of data sources to over 100,000 diverse data sources across news, patents, research, and government data types
    • Developed several Flask APIs served using Gunicorn for consumption of DS services by the front-end.
    • Built tools for client engagements—ontology creation best practices, automated ontology validity assessment, bash scripts with parameterization for custom app scheduling, and a fail-fast initialization scan to ensure all required configuration parameters are present
  • Post-Undergraduate

  • 2017 - 2019

    International business machines

    Senior Artificial Intelligence Engineer
    • In support of a Watson Explorer (WEX) client engagement, developed a batch document translation application utilizing Watson Language Translation from IBM Cloud Services to translate foreign language documents before crawling into the system. Additionally, the application could train custom language models using several different artifacts; a forced glossary, a parallel corpus, or a monolingual corpus
    • Designed and developed a Watson Conversation Service (WCS) demo that utilized several different API’s to provide historical weather information and named storm information. The integration layer was a custom built Node.js application that contained sophisticated geo-location processing that converted natural language locations to a bounding box of coordinates which were used to determine the closest weather station
    • Utilized the Cognitive Avatar as an interface for a WCS chatbot in an innovative engagement with a large financial services company. A custom Speech-to-Text model was built to capture the unique terminology of the client’s domain. Informational cards were also integrated to highlight important information for the user
    • Extended the capabilities of the Cognitive Avatar by implementing C# code that could capture a user’s mood and appropriately respond based on the detected emotion
    • Created several golden images of various Watson products to reduce deployment time for client engagements
    • Published defensive patents and publications outlining our existing work and new, innovative ideas
    • Worked with the business development team to provide technical guidance, gather requirements, and solution recommendations
    • Created several other client demos of WDA, WCS, and WEA
  • 2015 - 2017

    International business machines

    Artificial Intelligence Engineer
    • Developed a domain adaptation Java tool that aided in the creation of question-answer pairs for the Watson Discovery Advisor (WDA) application, IBM’s deep question-answer system originating from the Jeopardy! Challenge. The tool automatically generated standard variations for a wide variety of answer types allowing a domain adaptation specialist to work more productively—increasing consistency and accuracy. In effect increasing accuracy of the system
    • Performed end-to-end support for the WDA system involving developing client entity and lexicon artifacts, developing question-answer pairs, performing data assessments, ingesting documents into the system, Knowledge Graph loads, domain adaptation, developing new features, logging defects, and writing up documentation
    • Developed a data cleansing pipeline tool that was used for data assessments and cleaning client data for various engagements. The tool scrubbed documents of several common issues (difficult characters, excessive whitespace, gibberish, etc.) and then converted documents into a form acceptable for WDA ingestion
    • Completed a first-of-a-kind implementation of Watson Engagement Advisor (WEA) 3.0, designing and developing a conversational chatbot that connected back to the client’s customer management system allowing a user to retrieve personal billing information, request a billing extension, pay their bill and get general knowledge questions answered
    • Created an automated dialog testing tool for the Watson Dialog Cloud Service which read from a spreadsheet and simulated the conversational flows to verify the accuracy of the Natural Language Classifier service as well as confirm appropriate responses
    • Published defensive patents and publications outlining our existing work and new, innovative ideas
    • Created several other client demos of WDA, WCS, and WEA
  • 2014 - 2015

    International business machines

    IT Techincal Consultant
    • Designed, developed, and deployed a Java portlet that tracked the status of UC-2’s and UC-2A’s within the UCMS system
    • Managed a small team who documented the webMethods component of UCMS
    • Contributed to the creation of over forty manual regression test scripts for the Test for Production environment
    • Created a Maven version of each UCMS portlet project, and created unit test classes with JUnit in a shift to continuous integration for improved testability
    • Developed a Java application for ClearCase which reformatted the exported XML to be easily imported into Microsoft Excel for analysis
    • Identified, investigated and fixed Java defects within the UCMS system
  • Undergraduate

  • 2009 - 2013

    University of Maryland

    Bachelors of Science in Computer Engineering
    • Minor in International Engineering
    • Study Abroad in China – International Engineering & Business Cultures
    • Study Abroad in New Zealand – Environmental Sustainability
  • 2012 - 2013

    T-rex consulting corporation

    Computer engineering internship
    • Developed an application necessary for the tracking and security of documents in a bid for a loan-scanning contract
    • Created web forms necessary to the digitalization of the hiring process for the HR department
    • Assisted the project manager in building requests for proposals (RFPs), gathering supplier quotes, and maintaining the scheduled timeline for a Department of Commerce contract
  • Summer 2012


    Summer Engineering Co-Op
    • Developed the automatic proximity probe testing application for the remaining seven product lines
    • Streamlined the eight applications into four, as well as maintained and upgraded database tables to better assimilate with future products
    • At the request of the CEO, created a daily email reporting services in C# featuring charts using Google’s charting API
  • Summer 2011


    Summer Engineering Co-Op
    • Successfully implemented an automatic proximity probe testing application that reduced testing time by over 300%. The application manipulated a digital micrometer stage and was written using LabVIEW
    • Performed signal analysis testing several sensor product lines ensuring no manufacturing defects
  • High School

  • 2005 - 2009

    Fairport High School

    • 4.0 GPA
    • Took 4 Advanced Placement & Syracuse University Classes

My Interests

I love to travel! I have been to six of the seven continents and hope to reach the last one, Antartica, soon. While traveling, I love to hike and take tons of photos. For me, photography is a creative release from the mostly left-brained software development world. My photography hobby eventually evolved into the side-business, Alex Tonetti Photography. The business perfectly complimented my career and my interest in entrepreneurship. Additionally, I am very passionate about finance and the financial independence community, music and biking to explore my city.

  • Music
  • Hiking
  • Photography
  • Finance
  • Traveling
  • Movies
  • Cycling
development awards
photography awards

Get in Touch

Author Image

Feel free To Contact

Interested in working together? Have a question for me? Wanna tell me your childhood best friend's name? Whatever it may be drop me a line using one of the below.