Hello! I am a Machine Learning Engineer currently working at The Browser Company. I primarily build cloud-based and on-device ML applications, focusing on LLMs, NLP and search.
Contact me at: info@mattmckenna.io
Current/Past Positions (see LinkedIn for longer history)
Staff Machine Learning Engineer - The Browser Company (20244 - current)
Fine-tuning, LoRA, function calling, retrieval & embeddings, semantic search, rerankers, and on-device inference.
Senior Machine Learning Engineer (NLP and LLMs) - Spotify (2020 - 2024)
Built, maintained, and improved machine learning (mostly NLP and NLU) models that power the Spotify Voice system and serve millions of users in real-time.
Data Scientist - Amazon (2017 - 2020)
I worked on many ML projects as part of the ML Solutions Lab at AWS, including NLP, time-series, classification, and more. I also spent time in Alexa working on NLP/NLU models.
Statistical Programmer - Harvard University (2013 - 2016)
I built statistical models to analyze clinical trail data for HIV research (see selected publications below).
My AWS ML Blog Posts
Kinect Energy uses Amazon SageMaker to Forecast energy prices with Machine Learning
Selected Publications
Mottini A, McKenna M, Doty C, Kim A (2019, December). “Developing Deep Learning Models for NLP-Labeling Tasks”. Paper presented at the Amazon Alexa NLU Conference, Seattle, WA.
Nordell AD, McKenna M, Borges ÁH, Duprez D, Neuhaus J, Neaton JD (2014). “Severity of cardiovascular disease outcomes among patients with HIV is related to markers of inflammation and coagulation.” J Am Heart Assoc. 2014 May 28;3(3):e000844.
Taiwo B, Hunt PW, Gandhi RT, Ellingson A, McKenna M, Jacobson JM, Gripshover B, Bosch RJ. “CD8+ T-cell activation in HIV-1-infected patients experiencing transient low-level viremia during antiretroviral therapy.” J Acquir Immune Defic Syndr. 2013 May 1;63(1):101-4.
Moscicki AB, Yao TJ, Ryder MI, Russell JS, Dominy SS, Patel K, McKenna M, Van Dyke RB, Seage GR, Hazra R; Shiboski. “The Burden of Oral Disease among Perinatally HIV-Infected and HIV-Exposed Uninfected Youth.” PLoS One 2016 Jun 14;11(6):e0156459