Justin Hou

Incoming Tech Investment Banking Intern - UBS

• Support senior bankers in the execution of M&A transactions, IPOs, and capital raises for semiconductor and AI infrastructure companies.
• Perform valuation analyses to assess strategic opportunities for clients in the technology sector.

Investment Analyst Intern - TSV Capital

• Conducted due diligence and benchmarking of 10+ AI startups against comparable public companies using revenue multiples, growth metrics, and market positioning contributing to investment recommendations.
• Produced industry analysis of 3.5 B TAM Embodied AI sector, mapping key industry players and emerging technologies across growth and value companies, and presented strategic market overview to firm partners.
• Coded tool to scrape Crunchbase and source founders of 20+ prospective AI portfolio companies.

Machine Learning Intern - Lockheed Martin Solar and Astrophysics Lab

Stared at the sun.

• Developed recurrent neural network to analyze polarity inversion line evolution in solar magnetogram sequence images, predicting active region eruptivity with a 7-10% improvement over baseline methods.
• Built a GUI application in Python for Stanford astrophysics research collaborators to view temporal evolution of sunspots and eruptivity events, improving data accessibility and interpretability.

CEO and Founder — Lucidity

Healing brains from brainrot.

• Developed and piloted a classroom management application in 50+ classes to intelligently block distracting sites with AI, encourage students to foster a healthier relationship with learning, and reward productivity.
• Consulted 50+ teachers across Los Altos Christian School, Gunn High School, and Applied Computing Foundation classrooms to understand their concerns about classroom focus and lesson retention.
• 2020 Conrad Challenge Finalist in Transforming Education through Technology.

President - Columbia Organization of Rising Entrepreneurs (CORE)

• Directing and emceeing CORE Demo Day, managing a $10K event budget to organize Columbia's largest undergraduate event, inviting DRF Founder Molly Fowler and Ramp CPO Geoff Charles, hosting startup pitch competition, and facilitating connections between 100+ alumni, founders, and club members.
• Managing club funds, supporting entrepreneurs through CORE Enterprises, bringing cool speakers.
• Managing the Lion Launch fellowship, mentoring a five-team cohort to develop startups and pitch to venture capital firms, enhancing entrepreneurial skills and funding readiness.

Director - Almaworks Accelerator

• Evaluating and sourcing 50+ early-stage startup pitches quarterly considering market capitalization, revenue metrics for traction, scalability, and team dynamics to select six high-growth ventures for the portfolio.
• Organizing weekly mentorship sessions culminating in a demo day, connecting 200+ founders, investors, and alumni, resulting in enhanced funding and networking opportunities at the Columbia Startup Lab.

Research Assistant - Creative Machines Lab @ Columbia University

I. I. Rabi Scholar Researcher

• Working under Prof. Hod Lipson on self-modeling robotic systems capable of reasoning about their own structure and behavior.
• Developing algorithms that enable robots to autonomously infer internal models of their morphology and dynamics through interaction and observation. Integrating LLM-based reasoning and physics-informed simulation to allow adaptive planning, self-repair, and goal-directed behavior. The project aims to advance embodied intelligence, bridging self-awareness, simulation-based learning, and real-world physical autonomy.

Research Assistant - Computational Morphogenomics Lab @ Columbia University

I. I. Rabi Scholar Researcher

• Implemented cross-attention transformer architecture with spatial and gene expression embeddings for tissue segmentation and classification, achieving a 0.82 F1 score while handling sparse gene expression profiles.
• Optimized graph convolutional network and PCA model for semantic segmentation of spatial transcriptomic cell images through cross-validating five key hyperparameters, improving F1 score from 0.72 to 0.79.
• Integrated closed-loop enhancement to models using clustering loss and statistical error analysis.

Invited Research Presenter — American Astronomical Society

Only high school research team to present at American Astronomical Society 239.

• Discovery of Debris Disks in Kepler Habitable Small Planet Candidates: Presented research on discovery of potentially habitable Kepler Object of Interest exoplanet, Kepler-69c, using infrared-excess detections from debris disks emission.
• Press Kit