Hi, I’m a Software Engineer in Los Angeles

I'm an experienced software engineer with a passion for financial markets and designing/building scalable and efficient systems. I love learning about the details of a system and how everything comes together, from the application level all the way down to the kernel and hardware. Currently using Machine Learning and Artificial Intelligence to solve complex problems in the stock market.

In my spare time I love being active by working out at the gym, playing basketball, exploring different cities and nature spots, and working on cars!

Stock Market Research on the go

Have you ever felt like the tools provided with stock market mobile apps don’t provide the information you need to make the trade? Worry no more, as I have created a stock research app that does that and uses the power of AI to help predict stock movements!

Checkout my stock research startup here:

Conquering Machine Learning

Designed for those who have computer science backgrounds and want to learn about AI and Machine Learning at a deep level, or for those who want to start building their own algorithms, follow along as I delve into both the core concepts of machine learning and the Python code used to build and train your own powerful ML algorithms.

This is 100% FREE to follow along!

Experience

  • Founded a stock market research web app (ReactJS and Next.js) and a mobile app (React Native) that currently has 200 active monthly users

    Fabricated powerful machine learning algorithms using recurrent neural network models (LSTM) with AWS SageMaker and PyTorch, to predict price movements with 60-70 percent success rate

    Integrated AlphaVantage financial RESTful API to get stock data with an average latency of 250 ms

    Using AWS cloud services such as Lambda, Cognito, and DynamoDB (NoSQL) for executing serverless API calls, authentication, and to design a scalable database that stores user and stock information (i.e price points, earnings history, balance sheets, etc)

    Created a REST API and GraphQL API endpoint in AWS (API Gateway) to serve as a gateway between the client and the database servers

    • Worked as a backend engineer in a high-paced, microservices/event-driven architecture to prevent customers from abusing the Amazon eCommerce platform in 200+ marketplaces worldwide

    • Led designing and building a scalable distributed caching service (Redis for AWS Elasticache) to suppress payment methods and reduce average checkout latency by 150 ms per request

    • Managed a team of junior engineers and led expansion of a large scale application to handle service requests from 10K TPS to170K TPS worldwide using Kubernetes (EKS) and AWS CDK (Infrastructure as code)

    • Troubleshooted services on linux based systems to identify memory bottlenecks and improve performance by 100 ms per request

    • Developed and maintained REST APIs (OAuth 2.0) to ingest and process data from data pipelines and reduce computation times by 40 percent

    • Integrated risk signals into large scale data pipelines to pinpoint abusive customers and reduce abuse by $1mm per year

    • Implemented RPCs for refreshing authentication tokens, resulting in average service latency decreases by 20 ms per request

    • Invented dynamic configuration deployment strategies to reduce service configuration deployment times by 95 percent worldwide

    • Applied a modified version of the Model View View-Model (MVVM) software architecture to a distributed system enterprise level application (C#/.NET) in order to reduce and limit class dependencies and create more modular and readable code

    • Designed and implemented classes to create a reusable and scalable network messaging system between distributed systems that increased throughput by a factor of 4

    • Implemented asynchronous programming and multithreading to upload and parse files for adjusting FPGA parameters, thus reducing program runtime by 2 sec

    • Reduced program run time by 80 percent by identifying and eliminating reads/writes to disk

    • Worked as a backend software engineer on a simulation system that sends out RF and other signals to test different entities

    • Designed and built puzzles and logic games through the use of microcontrollers for an escape room which has hosted over 45K customers in Baltimore city over two years.

    • Incorporated and programmed Arduino Uno circuit boards in C to create puzzles that are built on mechanisms such as magnetics, RFID, light sensors, and weight sensors.

    • Interfaced with corporate representation and pitched rooms to clients such as Morgan Stanley, AT&T, Exelon for team building activities in order to boost respective workplace morale and productivity.

    • Produced 4+ star ratings on Facebook and Google reviews with over 100 reviews on each platform.

Side Software Projects

Along with my main passions, I love working on side projects. From working in C++20 and using libraries like boost to create a chat messaging app, to using xamarin forms to create a cross platform mobile app for tracking your expenses and finances with Plaid’s integrated API, I always look for ways to learn and apply new knowledge to create products that can help people.

    • Created a Order Book for placing and keeping track of stock market trades

    • The user can place the following types of orders: Market, Limit, Stop, TrailingStop, FillOrKill, GoodTillCanceled

    • Used Boost to setup a connection between client and servers using Websockets (HTTP Upgrade)

    • Github

    • Cross platform mobile app that allows the user to connect to their financial accounts via Plaid API

    • The financial account from the plaid API response and stored in a NoSQL database

    • Dependency Injection (C#/.NET) is used to reduce coupling and providing a more modular code design

    • Github

    • A chat application where multiple users can join a single chat room and talk to one another

    • The design follows a client-server architecture, where the user is the client and every time a message is sent by the client, it is received in a buffer on the server side and processed to display to the rest of the users

    • I used the Boost libraries for all the Networking infrastructure (sockets, i/o context, etc)

    • Multithreading is implemented so that clients are executed on their own separate thread, while the main thread is consumed by the server

    • Github

Contact me

If you’d like to reach out and learn more about my experience or have any questions feel free to contact me!