Move37Labs

AI Strategy

Assist your organization in understanding how artificial intelligence can best serve your needs. We look at your specific goals and challenges, and pinpoint where AI could be useful in your value chain. We conduct diligence, competitive analysis, create a project plan, and set you on the path towards flawless execution.

Project Lead

We provide AI leadership on a part-time basis for your technical teams. This helps teams remain focused, provide insight to the C-Suite, proactively identify blockers, provide mentorship for junior devs/researchers, and help make pivot/stay-the-course decisions.

Research & Publishing

We translate your company's AI research project to a publication. This work includes planning studies, careful testing, packaging, and storytelling. Such publications can be essential for obtaining funding, increasing awareness, gaining customers, and hiring.

Hiring

We help your company build the right technical team by setting up hiring pipelines around technical knowledge, culture fit, and compatibility.

Nucleic acid sequence design
(Google & MIT)

Led a research collaboration between Google, MIT, and Move37 Labs to benchmark in-silico nucleic acid designers. We developed a state-of-the-art designer, open sourced our benchmark, accepted to the 2025 ICML workshop on Generative Biology.

Genomic Language Models for Marine Biodiversity
(New Atlantis Labs)

Functioned as the fractional head of data science for Deep Atlantis, and developed Genomic Language Model-based models for taxonomic classification of marine environmental DNA. Projects include novel ways of taxonomic classification of truly unseen marine environment DNA, and product workflows.

Vocal biomarkers for Depression
(Nurtr)

Developed algorithms for detecting perinatal depression from voice. Helped conduct analyses and engineering. Work is published on the bioarxiv, in review at npj Digital Medicine.

Fetal Heart abnormalities during pregnancy
(HeartSounds)

Developed a product to detect fetal heart abnormalities in early stage pregnancies using at-home ambulatory doppler ultrasound. Applied for government funding and provisional patents for this work. Conducted novel algorithm development using signal processing, neural networks, and LLMs.

AI Agents for creating Healthcare Apps
(Quome)

Technical advisor for Quome, a company developing AI Agents for generating secure healthcare apps.

Targeted drug design
(BioDavinci)

Technical advisor to BioDavinci on applications of ML to therapeutic design.

Next-Generation TPUs for Genomic Models
(Google, Deep Genomics)

As AI Engineering Lead at Deep Genomics, organized testing of Google's latest TPU hardware for training and evaluating next-generation RNA foundation models. (blog)

TensorFlow GAN
(Google)

Designed and authored TensorFlow GAN, the standard TensorFlow library for training and evaluating Generative Adversarial Networks (code, patent, blog, course)

Knowledge distillation for fast and accurate DNA sequence correction

NeurIPS workshop (Google, Verily)

Combined Compression and Feature Extraction Models for Storing and Analyzing Medical Videos

Patent (Verily)

NucleoBench: A Large-Scale Benchmark of Neural Nucleic Acid Design Algorithms

ICML Workshop (in review) (Google, MIT)

Reducing health anxiety in patients with inflammatory bowel disease using video testimonials: pilot assessment of a video intervention

JMIR (Google Japan, Kitasato U Hospital)

A prospective evaluation of AI-augmented epidemiology to forecast COVID-19 in the USA and Japan (blog)

npj Digital Medicine (Google Japan)

Does video compression affect CADe polyp detectors?

Gastrointestinal Endoscopy (Verily, Google)

Predicting the generalization of computer aided detection (CADe) models for colonoscopy

Spring Discover AI (Google Japan, Sapporo U, Kyushu U)

The Need for Medically Aware Video Compression in Gastroenterology

NeurIPS workshop (Verily, Google)

Trillsson: Distilled universal paralinguistic speech representations (blog)

Interspeech (Verily, Google)

Towards Learning a Universal Non-Semantic Representation of Speech (blog)

Interspeech (Google)

Voice biomarkers of perinatal depression: cross-sectional nationwide pilot study report

In review (Nurtr, Quome)

Methods and systems for implementing on-device non-semantic representation fine-tuning for speech classification

Patent (Google)

Method for detecting and classifying coughs or other non-semantic sounds using audio feature set learned from speech

Patent (Google)

An Improved Measure of Automatic Speech Recognition Performance in Clinical Settings

United European Gastroenterology Week (Verily)

Comparing supervised models and learned speech representations for classifying intelligibility of disordered speech on selected phrases

Interspeech (Google, Verily)

Clinical BERTScore: An Improved Measure of Automatic Speech Recognition Performance in Clinical Settings

ACL Workshop (Verily)

Automatic speech recognition system contextually biased for medical speech

Patent (Verily)

Personalizing ASR for dysarthric and accented speech with limited data (blog)

Interspeech (Google)