CV

ML & AI

Frameworks

Languages

Web

Infra & Tooling

Math & Theory

Rush Enterprises is a Fortune 500 company focusing on commercial truck maintenance, parts, leasing, and more. Developed and deployed several high-impact AI models, apps, and APIs improving business processes.

• Engineered a versatile AI assistant leveraging agent teams, RAG (hybrid vector / BM25 search), and MCP servers to provide domain-specific insights for business operations and automotive repair.
• Developed computer vision and ML models with segmentation and scene text recognition, improving efficiency for customer search at Rush Truck Centers.
• Designed, developed, and deployed multiple Streamlit apps for graph-based part search, anomaly detection, and general data analysis.

Advent International is a world top-10 private equity firm with over $100 billion in assets under management. Directly influenced million- and billion-dollar investment decisions through data-driven insights gained from the completion of over 100 data science projects.

• Using over 15 data sources, analyzed a wide variety of KPIs such as momentum, retention, churn, customer lifetime value, web traffic, and user engagement for hundreds of companies.
• Characterized sentiment around investment targets and competitors using NLP techniques, including text summarization, topic modeling, and sentiment analysis, with state-of-the-art models (at the time) such as BERT.
• Drove a portfolio company relocation initiative for 4 retail stores and expansion plans for 60+ stores, using geographic data to perform suitability analysis, create traffic heatmaps, and quantify total addressable market.
• Automated sourcing of thousands of potential investment targets, linking standard information (e.g. funding) with FTE growth, web traffic growth, social media growth, and online reviews.
• Aided portfolio company by creating, training, and validating AI models to recognize faces and identify attributes from photos using CNNs.
• Managed entire data lifecycle, including scraping, utilizing AWS, SQL, and Snowflake, and visualizing with Tableau, Python, and Excel.

Pi School's School of AI is a highly competitive program (~1% acceptance) that provides engineers €12.5K grants to receive mentoring while working on real-world AI projects. Used Bayesian modeling to improve medical diagnosis models by 36–52% across multiple metrics.

• Substantially improved the core business product: medical diagnosis models. Compared methods including gradient boosting, Bayesian models, Gaussian processes, and neural networks, improving accuracy of multiple models from a range of 61–77% to 83–93%.
• Upheld stakeholder interests and clearly communicated technical details while working remotely in a team and researching relevant new AI articles.
• Utilized causal inference frameworks such as CausalNex to inform Bayesian modeling.

Holy Grail is a startup backed by Y Combinator in batch summer 2019.

• Led development of a successful machine learning edge device prototype and shipped it to clients.
• While working remotely, implemented and deployed cutting edge probabilistic machine learning solutions.
• Managed and added features to both the React frontend and the backend of the customer-facing web app, including the PostgreSQL database.
• Updated REST API and managed distributed cloud services (AWS).

Through tutoring, raised a University of Pittsburgh student's grades at least a letter grade in two classes: Physics for Science and Engineering II, and Analytic Geometry and Calculus II.

• Applied knowledge of educational psychology and neuroscience to teaching, such as scaffolding, mnemonics, spaced repetition, and the testing effect.

Worked in the SMILE Lab under Aaron Batista, PhD. Collaborated with University of Pittsburgh and Carnegie Mellon University labs on brain-computer interface (BCI) research. Funded by the NSF GRFP - a highly competitive fellowship that provided $126K over three years.

• Used multiple engineering and machine learning techniques, such as dimensionality reduction, Kalman filtering, and regression, to process and analyze complex, high-dimensional neural data from electrode array-based BCI experiments.
• Spearheaded lab code refactoring effort, emphasizing reuse, consistent style, documentation, and organization.
• Effectively communicated complex ideas through data visualization, presenting at journal clubs, and teaching as a TA.

Rotated in the Marlene Cohen Lab at the Mellon Institute, in joint program between CMU and the University of Pittsburgh - the Center for the Neural Basis of Cognition (CNBC). The lab focused on vision, attention, population coding, and neuronal correlations.

• Implemented multivariate Gaussian fits for receptive fields in MATLAB.
• Mapped contrast, spatial frequency, and temporal frequency tuning in a visual cortex experiment.
• Performed cross-correlation analyses on neural spike train data with jitter correction and normalization.

• Assisted Dr. Linda Rinaman for the Introduction to Neuroscience class. Proctored exams, graded materials, ran review sessions, and answered questions.

Rotated in Dr. Aaron Batista's SMILE Lab focusing on sensory feedback, brain-computer interfaces, motor and somatosensory systems.

• Quantified the effect of vibrotactile feedback for assisting BCIs by analyzing hand reaching trajectories.

Cognitive, Computational and Systems Neuroscience Summer Undergraduate Research Experience (C-SURE).

• Full-time summer research in the Blazquez Lab via the C-SURE program.

• Conducted oculomotor-behavior research in the Blazquez Lab (Otolaryngology), contributing to studies of cerebellar-cortex neuron responses in the alert primate.

• Conducted psychology research in the Memory Lab on learning and misinformation effects.

Thesis: “Clustering of Preferred Directions During Brain-Computer Interface Usage” · Advisor: Aaron P. Batista · Defended Sep 2, 2016

Committee: Steven M. Chase, Brent Doiron, Neeraj Gandhi

Selected coursework: Neural Data Analysis (CMU 86-631); Neural Signal Processing (CMU 18-698)

journal

Chiang, C. D., Lewis, C. L., Wright, M. D. E., Agapova, S., Akers, B., Azad, T. D., Banerjee, K., Carrera, P., Chen, A., Chen, J., Chi, X., Chiou, J., et al. (2012). Learning chronobiology by improving Wikipedia. Journal of Biological Rhythms, 27, 333–336.

article

Chiou, J. (2011). Searching for Internal Models Through the Saccade Main Sequence. Washington University Undergraduate Research Digest, 6(2), Article 36.

thesis

Chiou, J. (2016). Clustering of Preferred Directions During Brain-Computer Interface Usage. Master's Thesis, University of Pittsburgh.

Showed that neurons' preferred directions cluster during BCI use but not during natural arm reaching, with a model relating modulation depth to BCI controllability.

thesis

Chiou, J. (2012). The Saccade Main Sequence Is Built by the Inverse Model of the Eye Movement. Senior Honors Thesis, Washington University in St. Louis.

conference

Alemayehu, B., Pavlovsky, N., Chiou, J., Tyler-Kabara, E., Hatsopoulos, N., Chase, S., & Batista, A. (2015). Eye-centered tuning is weak in dorsal premotor cortex when monkeys are not trained to fixate. Society for Neuroscience Annual Meeting, 46705.

conference

Blazquez, P. M., Chiou, J., Ren, L., & Yakusheva, T. (2011). Response of large granular layer interneurons in the ventral paraflocculus during oculomotor behaviors in the rhesus macaque. Society for Neuroscience Annual Meeting, 922.16/VV12.

with Lynn Ren, Eric Hsu, Mallory Highstein, Pablo Blazquez

with Rachel Diamond, Pablo Blazquez

with Franklin Zaromb, Henry Roediger III

Prototyped Dagu, an app concept providing pastoral communities with networking capabilities and information about water availability, grassland, market prices, and safe routes.

$126K over three years. Designed a brain-to-brain interface experiment in the 2012 proposal, predating the first paper published by another group (Pais-Vieira, Lebedev, Kunicki, Wang, & Nicolelis 2013); also proposed intra-brain artificial circuits.

Received Dean's List award for multiple semesters.

Cognitive, Computational and Systems Neuroscience Summer Undergraduate Research Experience program.

• *99.9+ percentile rank by official ETS comparison to GMAT and LSAT. github.com/jessebmurray/GRE-percentiles
• **340 by concordance and modern percentiles.

See Experience section.

Taught high school students about brain-computer interfaces and neuroscience as part of CNUP's outreach program.

Tutored 5th- and 6th-graders weekly in math, English, science, and history for the full academic year.

Supported teachers in caring for and engaging neurodivergent children ages 2–5, including many on the autism spectrum.