Relevant Coursework: Probabilistic Models and Machine Learning, High Performance Machine Learning, Biostatistics for Engineers, Advanced Programming, Data Structure, High Dimensional Stats for Biomedical Data, Deep Learning for Biomedical Imaging, Quantitative Physiology, Calculus III, Linear Algebra

• Developed discrete diffusion model for protein sequence-structure co-generation using ESM3 structure tokenizer and BERT architectures, achieving faster generation speeds compared to continuous diffusion approaches.
• Implemented sparse autoencoder training pipelines to uncover interpretable biological features in protein language models. Work was published at ICML 2025 with highlights.
• Designing algorithms to select a maximally diversified set of proteins for molecular dynamics simulations, generating data to support subsequent model development for predicting protein conformation trajectories.
• Developing methods to extract protein conformational ensembles from AlphaFold2 through latent space exploration and systematically created a benchmark library to compare with existing methods.

• Co-led development of DNA Discrete Diffusion (D3), a score-entropy discrete diffusion model for conditional generation of regulatory sequences. Implemented variant effect prediction extensions and built latent space visualizer to interpret learned regulatory motifs.
• Established fine-tuning pipelines with Low-Rank Adaptation (LoRA) and Supervised Fine-Tuning (SFT) for four pre-trained DNA language models to evaluate its representational power for regulatory genomics.
• Developed and implemented evolution-inspired data augmentations (EvoAug-TF) in TensorFlow for genomic deep neural networks and demonstrated its improvement in generalization and interpretability.
• Designed and evaluated more than 100 deep-learning models for predicting DNA promoters' expression rates using Python, TensorFlow, and WandB in the 2022 DREAM Challenge. Placed 7th in the final leaderboard.

• Developing a multimodal transformer model to predict binding affinities of protein-molecule bindings.
• Optimized the training and inference pipeline of protein language models by up to 30% by reimplementing models using flash attention and provided a validation method to evaluate data from wet-lab experiments.

• Developed CICaidA, a centralized health monitoring system for nursing homes featuring custom ESP32 hardware with MAX3010 sensors to track heart rate and blood oxygen, a Flutter mobile app, and Firebase backend to provide real-time alerts to caretakers.

• Built an MCP server connecting Zotero research libraries with AI assistants (Claude, ChatGPT, Cursor) via the Model Context Protocol. Implemented semantic search with multiple embedding models, PDF annotation extraction, and BibTeX export. Achieved over 720,000 downloads and 3k+ GitHub stars.

• Developed a protein structure visualization tool using the Molstar framework with support for multiple formats (PDB, mmCIF, MOL2, etc.), diffusion animation playback, sequence alignment, and ESMFold integration. Released as a VSCode extension (1100+ downloads), web application, and Jupyter notebook plugin (2000+ downloads).

• Developed deep learning models to predict small molecule-protein interactions using the Big Encoded Library for Chemical Assessment. Implemented over 40 types of DL models, including CNNs, GNNs, Transformers, RNNs, and GDBT Models, and finally developed a robust solution after using up all 480 submissions.

• Organized annual Biotech Summit with over 200 participants and hosted 8+ speaker events with guests from academia and industry with total attendance exceeding 400 people.
• Started data science competition initiative to help students learn AI through hands-on work with biological data; led a team of four students to finish top 5% (silver medal) in a bioimaging competition.
• Coordinated multiple initiatives including an iGEM team, a podcast series featuring professors and industry leaders, and outreach workshops introducing high school students to computational biology.

• Organizing and facilitating a series of biotech & deeptech speaker events and investor dinners, connecting 15+ industry professionals with Columbia student entrepreneurs, resulting in mentorship opportunities and potential funding partnerships for early-stage biotech ventures.

• Hosted biweekly machine learning/python hands-on workshops to cultivate members in their interests in AI.
• Collaborated with AI professors to create research opportunities and organized guest lectures from industry professionals.