Joel Bartlett Email & Phone Number

New York, Ny, Us

San Francisco, California, Us

San Francisco, California, Us

For an internal hackathon, I developed and pitched an Advanced Retrieval-Augmented-Generation (RAG) service using unsupervised learning algorithms.The idea was allocated engineering resources, and I then developed, productionized, and internally marketed the project. After conducting user-facing A/B Experiments, it is now rolled out and acts as a smart pre-processing layer for all LLM Summarization Requests at Slack AI.

San Francisco, California, Us

I was on Slack's Build Team - an Internal Tooling team responsible for CI/CD systems company wide. Along with maintaining and monitoring Jenkins and Github Enterprise at Slack, I migrated multiple Jenkins Instances from EC2 based controllers (provisioned via Chef and Terraform) to Pod based controllers (using Docker, Kubernetes). This work to modernize Slack's Jenkins controllers provided significant business-level benefits, including:1. Resolving security vulnerabilities automatically by recycling controllers with the latest stable releases.2. Reducing cycle times on controller hosts from literally 2-3 weeks to 2-3 minutes (this saving substantial costs in engineering hours)3. Replacing outdated processes for upgrading host dependencies and packages with efficient, robust methods - that not only save many engineering hours now but also prevent the multitude of incidents and SLA outages we saw before.

San Francisco, California, Us

Seattle, Wa, Us

After graduating college, I was blessed with a little time off to pursue a lifelong dream and work as a barista! I not only sharpened my customer service abilities and communication skills, but I also mastered the extensive Starbucks menu and drink recipes.Moreover, I gratefully received shout outs from customers who messaged the district manager, saying: "[Joel] was so kind and friendly while waiting for their drive thru order. His kindness is definitely something we need more of in this world. From now on they're only coming to this Starbucks for Joel."

Atlanta, Ga, Us

As a paid, undergraduate researcher at the Healthcare Innovations and Translational Informatics (HITI) lab, I developed a Deep Learning model using Stacked, Bidirectional LSTMs to predict if a patient is at risk of hospitalization within a 30 day forecast.

San Francisco, California, Us

During my Summer Internship I created a Slack Bot app to identify previous Incidents similar to the current Incident. For this project, I applied Cosine Similarity functional to Document Embeddings of Slack “Channels” to perform semantic search. Moreover, I participated in weekly ML Paper Reading Group discussions within the team, and even progressed in my specialized understanding of Machine Learning applications in Natural Language Processing to lead the final Reading Group session of the Summer. I presented a paper on “Unintended Memorization” of sensitive personal information in training models on large corpuses of text data. However, rather than walking through the paper alone, I went even further and used personal time to develop an implementation of the models and algorithms described in the paper.

San Jose, California , Us

After purchasing the components, I assembled a ML server with GeForce GTX 1080 GPU and set up the Ubuntu environment with all necessary drivers to train Deep Learning models. Using low-level Tensorflow APIs, I then created a DC GAN and later a Boundary Equilibrium GAN (BEGAN) based on published research. The objective was to generate synthetic facial image data for other teams to train a self-checkout kiosk classifier. I monitored training with Tensorboard to debug issues and tune hyperparameters.

Atlanta, Georgia , Us

As an undergraduate researcher, I collected experimental data of the human gait cycle to improve intent recognition classifiers of our lower-limb, active prosthetic device. In addition, I used Autodesk to calculate the Moment of Inertia Tensors of our active prosthesis and a passive market prosthesis to generate realistic simulations of the devices. Furthermore, I spent extra time after completing assigned tasks learning the basics of hyper parameter tuning and feature extraction from graduate students.http://www.vip.gatech.edu/teams/robotic-human-augmentation

San Jose, Ca, Us

As FRC Captain, I led the mechanical design of our world champion robot and taught newer members Solidworks and design principles. I also led fabrication and assembly of our robot and was a key member in our Pit Crew during tournaments.

San Jose, Ca, Us

Bachelor Of Science - Bs, Computer Science

Education record