Hi, I'm Jose
I'm an Engineer at Apple working on next-generation watch and audio products. My focus is on backend, internal tools, and IoT technology.
Since November of last year, I've been working on this API tailored for searching and tagging images sourced from Drakes menswear magazine and clothing store, ultimately serving them in JSON format. The purpose is to provide a search tool for outfit inspiration from their tailored looks and style. To manage the whole operation on my own, the development pipeline is streamlined with end-to-end continuous integration and deployment practices, which encompass automated testing and adherence to PEP 8 standards through Github Actions and automatic building/deployment and management of Docker containers with Google Cloud Build/Run services. I scrape all images to populate a CDN with the latest photo shoot images from Drakes. Image links are efficiently indexed and tagged within MongoDB, with hosting using Mongo Atlas servers. (P.S. I used the API for the image in this section)
The inspiration for this app is the feeling when you spontaneously see a friend at a bar or coffee shop without planning anything in advance. The goal is to encourage people to share their in-person activities with friends by posting a pin on the shared map. Others can see the map in their free time and choose to drop in without further back and forth planning. Its backend is built with FastAPI and MySQL, using JWT tokens for authentication and location mapping. The frontend is built with React, and Bootstrap for styling with a clean user interface built on top of Mapbox allowing users to see their friend’s activities on the map and join their group.
I'm part of the Human Interfacing Devices team at Apple where I work on Watch, Airpods, Beats, and future products. My main role is to support algorithm development, firmware development and user experience studies with everything they need by automating data gathering and bring-up from sensors on prototype devices. I also plan and demo sensor features and automation strategy at new product and status reviews. Within my team, my main focus is automation efforts on ‘in-ear detection’ for audio products and the Watch touch technology, but I also support other features such as Airpod's stem UI and water submersion detection for watch.
This app uses Assembly AI's speech-to-text API to process podcasts (or any audio) and summarizes them into usable prompts for the stable diffusion art generator. It can convert any audio smaller than 200mb into a paragraph summary or intelligently break it into chapters. I use these as prompts for Stable Diffusion based API that produces amazing AI art! Available here or on my Github, this app’s front end uses Streamlit for clean UX and easy deployment.
The Boozeboy smart speed-rail is an IoT inventory system that monitors real time changes in the volumes of all the bottles currently being used in a bar or restaurant environment. My team also implemented a data visualization platform using Losant, that allows bar managers to see their real-time inventory needs and gain insights based on their long term customer consumption. Booze boy allows bar owners to track how much alcohol is going in each drink, and at what time without relying on receipts. The data is captured using force sensors embedded within our specially designed speed-rails and then pushed to the cloud through MQTT. This data is then displayed to bar owners on their specialized dashboard where they can see a real time view of what’s happening in their bar.
By recomendation from the Electrical Engineering Department at FIU, Eco-tabs hired my partner and I on a contract to design and build an automated dispenser for bio-waste treatment. The dispenser was needed to treat hazardous H2S gas from waste in airports where sending personnel was dangerous. We worked with an industrial designer and machine shops to create multiple real-life prototypes to test on-site in toxic environments around the world. The main obstacles in this project were creating an electrical and software system with extreme power efficiency to reduce battery swaps(hazardous and expensive to perform) and creating a mechanical system that worked without fail in various harsh environments. In this project our two-person team completed all electrical and software features from the PCB to Embedded Software ourselves and worked closely with a designer to finalize the mechanical system and casing.
Our 2-man(and one fish) team won 1st place in hardware hack and 3rd place overall at the largest hackathon in Florida. We used an embedded microprocessor and infrared sensors to monitor a fish’s movement within a small tank and actuate motors accordingly to move the tank in the direction the fish swims. The infrared sensors were set up like a cage, about a half inch from the walls of the fish tank. When the fish aproaches a wall of the tank, it trips the sensor which signals the processor to actuate the motors in that direction.
During my internship program at GE Appliances, I had the opportunity to work with a team of other interns for a company hackathon. We decided to build an in-fridge rapid cooling system that focused on getting standard bottles and cans from room temperature to fridge temperature in 60 seconds. Using a DC motor and a hack to the fridge's ice and water dispenser, we sprayed ice cold water on the surface of the vessel while spinning it to bring the liquid in the center(hottest part) towards the surface. This movement allowed us to bring the can towards fridge temeperatures faster than regular air or water exposure could. In this project, I led the control systems and UI of the device.