Machine Learning Engineer

REHAN CHINOY

************@*****.*** 619-***-****

linkedin.com/in/rehanbchinoy github.com/rehanbchinoy SUMMARY

Machine learning engineer with a robust foundation in deep learning, distributed computing, and ultra-large databases. Proven track record in developing high-impact models and algorithms for drug discovery. Seeking opportunities to apply expertise in ML and data science to solving complex challenges. EDUCATION & ACADEMIC HONORS

University of California, Los Angeles September 2018 - June 2022 B.S. Applied Mathematics, Neuroscience Minor GPA: 3.85 Awards and Honors: Dean’s Prize for Excellence in Research, Undergraduate Research Scholarship Recipient Relevant Coursework: Machine Learning, Optimization, Advanced Linear Algebra, Advanced Differential Equations, Advanced Probability and Statistics, Numerical Methods, Data Structures and Algorithms EXPERIENCE

Frontier Medicines December 2022 – February 2024

Machine Learning Engineer

• Spearheaded the development of a bespoke search algorithm for the small molecule virtual space, resulting in an approximately 10% increase in hit rates across 5 pre-clinical projects.

• Engineered chemical property models utilizing ensemble architecture and transfer learning techniques, increasing prediction accuracy by 15% and rapidly accelerating drug discovery timelines.

• Led the deployment of models using MLFlow and serverless inference endpoints, reducing time to deployment into the drug development pipeline from days to minutes.

• Established and trained a proprietary large-language model as a “foundation” molecular embedding model, enhancing the organization's capabilities in molecular representation learning and increasing downstream model performance by 8%.

• Orchestrated computational workloads on distributed clusters using Spark and Sagemaker Pipelines, resulting in a 90% reduction in runtime cost.

sFOX August 2022 – October 2022

Software Engineer Intern

• Enhanced users' digital wallet security by implementing a state-of-the-art multi-party computation (MPC) security layer, bolstering protection against unauthorized access.

• Containerized the security layer using Docker and Kubernetes, facilitating seamless deployment to production environments.

Buonomano Lab, UCLA October 2019 – June 2022

Computational Neuroscience Research Assistant

• Published a computational neuroscience paper on interval timing mechanisms, serving as the first author in a peer- reviewed journal publication.

• Developed a biologically-inspired recurrent neural network (RNN) in Python using TensorFlow, utilized for modeling timing and working memory tasks.

• Leveraged MATLAB for in-depth analysis of network dynamics and generation of data-driven insights into neural representations of timing and working memory.

PROJECT HIGHLIGHTS

• MolSearch: used transformer-generated embeddings of drug-like molecules for billions-scale molecular similarity search

• Graph-based property prediction: used variations of GNNs for molecular property prediction

• PocketScout: used a geometric vector perceptron (GVP) to predict binding pockets on hard-to-drug proteins

• Interval Discrimination: trained a biologically-inspired RNN to discriminate between two auditory intervals, culminating in a first-author publication

• Literature Generation: trained a custom RNN (LSTM, GRU) and transformer to produce fake Homer and James Joyce text, demonstrating proficiency in NLP techniques TECHNICAL SKILLS

• Programming Languages: Python, SQL, C/C++, MATLAB, Julia, R

• Machine learning using TensorFlow, PyTorch, SKLearn, Keras, Jax

• Distributed computing using PySpark, Spark, Dask,

• ML-Ops using MLFlow, Lambda, ECS, Docker, Kubernetes, Airflow

• SQL/No-SQL database and data warehouse development using Postgres, MongoDB, Iceberg, Redshift

Contact this candidate