Alex Gracia
My mission is to transform ideas into reality with the use of technology. Blockchain, AI and Web Development are some of the technologies that I dominate and use in my projects. I use my creative mind to break down ideas and build a product that meets client expectations. I believe these disruptive technologies allow us to solve and improve our challenges, therefore there’s room for further developments and I’m looking forward to more projects to be of greater impact.
Experience
Software Engineer
- Worked with the Development team in the AspenNft platform as a Full Stack Engineer. Leveraged Typescript for coding, employed React.js for Frontend development, utilized Node.js and PostgreSQL for Backend development, and utilized Solidity for Smart Contracts.
- Contributed to the development of the minting engine responsible for creating NFT collections with both ERC-721 and ERC-1155 standards under Ethereum Virtual Machine (EVM) chains.
- Played a pivotal role in the development of a blockchain-based subscription engine. This feature assisted creators in generating revenue through a subscription system and helped us amass our initial customer base.
- Developed a redemption system that allowed clients to exchange NFTs for physical objects.
Full Stack Engineer
- Main full stack developer of Nave.finance , DeFi platform for investing using copy-trading in the Binance Smart Chain (EVM chain).
- Used Typescript as a language, React.js and Next.js as a Frontend, utilized Node.js and Mongodb as a backend development. and Solidity for Smart Contracts.
- Created the end-to-end flow for building an investment portfolio, showing the investments in responsive Cards and Tables and creating the discover page to find the portfolios. Developing in the front-end (components & pages) back-end (API & contract interactions).
Machine Learning Engineer
- Co-led the development of a Darkstore project. Which envolves technologies such as: object-detection, speech recognition and tracking . Also connects "things" with a backoffice developed in Javascript and Node.js
- Led the development of Covid Defender project. ML solution to detect mask, temperature, capacity and social distancing with Tensorflow.js, Node.js and React.js
- Created a end-to-end solution to connect an Avatar (Metahuman of Unreal Engine) with a Backoffice developed in Node.js and Express.js
- Developed a Dynamic Pricing solution connecting StockX with Lenovo screens showing the price dynamically via a Backoffice.
Machine Learning Engineer
- Developed for Autel Robotics EVO drone project (Top 10 Drone company worldwide) under Prof. Fernando de La Torre and Dong Huang.
- Developed a lightweight object detection framework for embedded devices in PyTorch, which connected feature extractors with detection heads, including lightweight Computer Vision (CV) algorithms to improve performance.
- Transferred models from Pytorch to Tensorflow via ONNX framework. This made improve the performance allowing to use TensorRT.
- Optimized potential models by transferring implementation to TensorRT, significantly improved the speed of object detection from 5 FPS to 17 FPS for NVIDIA Jetson TX2 embedded device.
- Established testing metrics (including mAP, TOP-F1, speed and size) to evaluate the suitability of models for further deployment in the drone.
Software Developer
- Created a Data Engineering from two clothes companies in a real environment. Applying MySQL queries to extract the data of the fashion retail shops.
- Created parsing scripts using Pandas and Numpy to label, clean and prepare the data for further training ML models.
- Applied Feature Engineering using libraries such as Pandas and Numpy using KPIs of the retail sector.
- Applied ML algorithms and tecniques to deal with missing data and normalize the features.
- Applied metrics to evaluate model performance (AUC, F1, Precision, Recall). Also, visualized data with tools like Plotly and Matplotlib.
- Created a predictive system to evaluate customer purchasing decisions based on transaction information.
Education
Carnegie Mellon Univeristy
- 10-601 Machine Learning, Prof. Matt Gormley.
- 10-707 Deep Learning, Prof. Russ Salakhutdinov.
- 16-720 Computer Vision, Prof. Simon Lucey.
Ramon Llull University
- Major in Algorithms and Image Processing.
- Student representative.
Portfolio
Aspen
💡 Idea:Redefine community membership in web3, set up crypto subscriptions, recover lost royalties and set on-chain IP protection.
💻 Technology: Solidity, Typescript, React.js, Node.js, Blockchain concepts, PostgreSQL.
✅ Reality:Created a platform with features to help creators such as: create NFT collection, subscription engine, recover royalties, redemption program, minting engine.
Currently published on production. See in this link.
Blockchain exchange project
💡 Idea: Create a blockchain exchange and a ERC-20 token.
💻 Technology: Solidity, Javascript, HTML5, React.js, Node.js, Blockchain concepts, Ganache/Truffle.
✅ Reality: Exchange that lets you deposit and withdraw ALEX token, token created under ERC-20 protocol and deployed in the Ropsten Test Network. The functions of this exchange are: to buy, to sell, to cancel orders and make ALEX/ETH & ETH/ALEX swaps. On the other hand, it's possible to see the historic orders made in the exchange and a chart of how the price of ALEX/ETH in time.
Currently published on production. See in this link.
Darkstore
💡 Idea: Create the concept store for the future of retail and IoT.
💻 Technology: Node.js, Javascript, React.js, Unreal Engine, Speech Recognition, Object Detection, Tracking.
✅ Reality: Created a shop from scratch applying state-of-the-art techonologies which auto-detects the people who enters in the shop with the Avatar interaction. With the conversation and insights of the avatar, the shop auto-adapts to the client profile offering him a products that are likely to buy. At the same time, our system gets analytics and data of the interactions that the client does in order to improve the conversion for future clients.
Check shop link.
Machine Learning algorithm for a drone company
💡 Idea: Real-time object detection algorithm for a drone company.
💻 Technology: Python, Pytorch, TensorRT, Tensorflow, OpenCV and CUDA.
✅ Reality: Fully functional algorithm for a drone that detects persons, cars, animals and riders in real-time. Also uses state-of-the-art ML models optimized to save battery comsumption fo the drone.
Check company's drone model where the algorithm has been developed for in this link.
Covid Defender
💡 Idea: Solution to make the places more safe due to COVID-19.
💻 Technology: Python, Tensorflow, Tensorflow.js, TensorRT, JavaScript, HTML5, Node.js and AWS.
✅ Reality:Product that combines that detects mask, temperature, affluence and social distancing with the power of IoT and AI technology. It helps to manage and control access to retail, hospitality, entertainment, educational and commercial spaces to help you in keeping your customers/guests/visitors and staff as safe as possible. Currently deployed in Spain, Middle East and South-America.
Check Admira's section of the product. See in this link.
CRM SMS Integration
💡 Idea: Send SMS from Segmentail, company which uses Zoho CRM for fashion retail.
💻 Technology: Javacript, Deluge, HTML5, CSS, Esendex & Zoho API.
✅ Reality: Module inside Segmentail/Zoho CRM which lets you send SMS to clients with workflows and rules, e.g., send SMS to client X if is the date of the birthday & bought 500€ in shoes. On the other hand, lets you check the historic of SMS that yous sent to the client. An advantage of the app is that it's desployed Zoho servers, this means that it's seamless and does not needed extra costs on external servers.
Check Segmentail company where the development has been done. See in this link.
Real State website for South Ibiza Villas
💡 Idea: Website for real state company
💻 Technology: Javacript, Css, HTML5, Bootstrap.
✅ Reality: Real-state website that lets you make reservations, get analytics, contact customers, make forms and send automatic e-mails to the customers. Also, has a map to check the properties that are available and contact via Whatspp to the real-state agent.
Check South Ibiza Villas website in this link.
Skills
- Programming Lenguages: Python, JavaScript, Solidity, HTML5 and Bash
- Web Frameworks: React.js, Node.js and WebRTC.
- Deep Learning Frameworks: PyTorch, Tensorflow, Tensorflow.js, Caffe and TensorRT
- Data Science Tools: Numpy, Pandas, Scipy, Seaborn and Matplotlib
- Databasess: MongoDB and MySQL
- Natural Languages:Fluent Catalan, Spanish and English
Interests
Apart from being an Engineer, I enjoy most of my time being outdoors. I participated in a half-marathon back when I was living in Pittsburgh. Also, I am passionate about martial arts, concretely Muay Thai which I practice regularly.
Music is one of my ways to disconnect. I started La Cúpula Barcelona with friends, a local music event company. Also, I enjoy creating content to share my knowledge altruisticly to people interested in technology (tech.with.alex).
To add, one of my characteristics is being highly curious. I spend many hours reading books of topics I am interested at that moment and join a variety of bootcamps and coursers to keep feeding my intelectual interest.
Awards & Certifications
- Google Analytics Certified Developer
- Mobile Web Specialist - Google Certification
- 1st Place - University of Colorado Boulder - Emerging Tech Competition 2009
- 1st Place - University of Colorado Boulder - Adobe Creative Jam 2008 (UI Design Category)
- 2nd Place - University of Colorado Boulder - Emerging Tech Competition 2008
- 1st Place - James Buchanan High School - Hackathon 2006
- 3rd Place - James Buchanan High School - Hackathon 2005