cesar juarez

a) Designed and developed of a series of Recommender models that suggest carriers to customer representatives. After creating a new model based on Learning to Rank the Top 10 metric increased from 20% to 45%.

b) Hackaton project. Recommender system that suggests carriers for shipper contracts with Drop Rate and OTD optimization.

c) Implemented an automatized and reusable pipeline service for fetching, transforming data and training Machine Learning models. Alongside, I developed a Machine Learning Server that supports loading and deploying multiple ML models in parallel.

d) Designed and launched a Drop Rate model that predicts the probability of a shipment to drop. The model was tested in experimental settings (A/B test) in order to decrement the Drop rate (5%) and improve the OTD.

e) Designed a Feature Store that provides historical features for our models and other downstream tasks.f) Development of a Machine Learning Platform that accelerates experimentation and placing into production Machine Learning pipelines for all the teams in the Data division.

g) Tech talks. - Learning to Rank for Recommender Systems. - Introduction to A/B tests.

Responsibilities: Natural Language Processing and Machine Learning Lead.

a) Developed data ingestion engine that that downloads royalty free english clips, cleans and transforms the embedded text to be used by our ML models.

b) Implemented a Recommender System that matches content (Movies, Series, Songs) to the current user’s preferences and performance. The system is fed with the user’s social networks activity (Facebook, YouTube and Spotify) to improve the model. Using a "Question engine" we create personalized questions that are presented to the users when they enroll in a course.

c) Developed ML model that classifies videos by content rate and filters content to users.

Technologies used: -Cloud platform: GCP -Back end: Python, Scrapy, Beautiful Soup, Spacy, NLTK, Pandas, Flask, MLFlow -DevOps: Kubernetes, Docker -Data orchestration: Airflow, Dataproc, PubSub, Kafka, Pyspark -Databases: Firebase, Neo4J -ML Frameworks: Xgboost, SKLearn

Responsibilities: Full stack (Front End and Back End) developer using J2EE technology.

a) Worked in the development and integration of a new call center and an ERP for CONSULMED.

b) Integrated a new platform that supports CONAGRA’s operations, sales and logistics for small shops, mom and pop stores. Jointly I developed the synchronization of the system with CONAGRA’s local ERP and BPCS.

c) Developed from scratch the meet-medic medical platform. It is an ERP where users (doctor or administrator): create medical appointments, administrate doctors and patients, send push messages to patients, have a payments ledger, create electronic bills. d) Developed a hybrid mobile app where the patient can schedule appointments, receive messages from doctors, look for hospitals and clinics that are registered in the platform.

Technologies used: -Front end: Primefaces, Jquery, Bootstrap, Angular-JS, Angular 2, HTML, Ionic 2 -Back end: Kafka, Enterprise Java Beans, Java Server Message API, Spring Boot (basics), REST web services, with JAX-RS, JPA. -Data bases: PostgreSQL, SQL Server.