About

In 2019, at my inaugural FiveStars hackathon, I developed a Slack bot that efficiently addressed the top 10 questions commonly posed by product and QA teams. This innovative solution earned me the first prize.

As a newcomer to the company, my small team and I were responsible for a microservice that authorized live transactions. My primary focus was on integrations with external services like WorldPay and FiServ, as well as ensuring accurate transaction status tracking across all possible outcomes.

Given the extensive testing in various environments, product managers and QA personnel frequently asked me similar questions, such as: "Did the transaction using POS go through? What’s the status in the database? What are the transaction details (amount, etc.)?"

To address this issue, I compiled a list of the top 10 questions and created a Slack bot that allowed product and QA teams to access the information they needed through a self-service approach, without requiring developer assistance. Since the teams were already using Slack for regular communication, this solution was a seamless addition.

In 2018, at my inaugural TrendMicro hackathon, I collaborated with a team of four to develop an AI/ML-based bot capable of playing poker. The challenge of learning numerous new concepts in the ML field while applying my expertise in creating Python-based microservices and deploying with Docker was invigorating.

The initial phase involved learning about Monte Carlo simulations and developing a fundamental static decision set for a given hand based on probability. One of my team members took the lead in setting this up.

Once the model was established, I employed Python and Flask to expose the model as an API, enabling it to participate in the contest. To ensure consistent deployment, I created a docker-compose file, allowing all services to run seamlessly with a single command. This approach facilitated the effortless deployment of various bots as needed.

Our preliminary version performed admirably among the bots in the Austin office. However, we aimed to enhance its capabilities further. I dedicated a significant amount of time to exploring various methods to fine-tune the model and test different iterations, ultimately identifying an optimal representative bot for our team.

Despite being one of the most challenging hackathon projects I have ever undertaken, the opportunity to expand my knowledge in ML topics proved to be invaluable.

In 2012, my team secured first place in 'Sprint of Dreams,' an internal hackathon at NetSuite. As part of a large group, I served as a developer and contributed to the implementation of some features.

The contest began with numerous promising ideas, and we had to vote on the one to pursue. This was an excellent team-building exercise, as we had all just met and discussing ideas served as an icebreaker.

Several factors contributed to our success:

In short, our winning formula involved solving real problems through collaborative effort and delivering a strong presentation. This experience was notably different from the one I had during the PayPal hackathon.

As a result of our victory, we all went home with brand-new iPads!

In 2008, PayPal initiated an internal hackathon competition for its employees called Labrats. At the time, I had just started my role as a Technical Product Manager (TPM) and didn’t give much thought to the contest. However, a friend who had submitted an idea and advanced through the initial round asked for my help in improving the presentation, particularly the business case and financial aspects.

Given my new TPM role, I gladly accepted the challenge. The idea aimed to enhance an internal tool used by developers and continuous integration systems to increase efficiency, saving both time and money. Our sales pitch revolved around the potential to save millions of dollars over several years.

We began by developing a model that considered factors such as:

We then created three scenarios—worst case, best case, and typical outcomes—if our idea were implemented. I focused on crafting concise PowerPoint slides that highlighted key points and showcased our model. Presenting remotely, I couldn’t gauge the judges' reactions, but one commented, "If I can save $9 million USD in 3 years, then I want it starting from right now!" This hinted that our final round performance was strong.

Following the presentations, the judges scrutinized our model extensively, questioning every estimate. It was a challenging period, but ultimately, our idea won second place!

The victory offered bragging rights for a year and boosted my confidence in presentation skills. So, when you have the opportunity to participate in a contest, give it your all—you never know what you might achieve!