Eliminating boilerplate infrastructure so startups can focus on what matters. Venture backed.

Grew company to product market fit, dozens of happy customers, and from a team of 2 to a star-studded cast of 20. Designed and led development of knowledge graph system, the data engine powering Harmonic.

15+ angel investments in technology companies

Worked on personalization (p13n) team to deploy recommender system machine learning models within the Play ecosystem to 1B+ users.

Worked within the Computer Vision Geometry Lab (CVGL), under the mentorship of Amir Zamir and Silvio Saverese. Research published at NeurIPS 2022 (https://taskdiscovery.epfl.ch/).Research topics include: computer vision, transfer learning, cross-task domain consistency enforcement, task discovery.

Fall 2016-2017 --- CS221, Artificial IntelligenceWinter 2016-2017 --- CS224N, Natural Language ProcessingSpring 2016-2017 --- CS224S, Spoken Language ProcessingFall 2018-2019 --- CS221, Artificial IntelligenceResponsibilities included: holding weekly office hours, grading assignments & final project reports, exam development, exam proctoring, leading section, etc.

• Successfully implemented an ensemble of recommender systems (content-based, collaborative-filtering-based, and heuristic-based) to match candidates with jobs, under a limited and noisy data setting.• Performed unsupervised learning (clustering) and supervised learning subtasks for prediction of: seniority level, functional area, functional titles, industry, and candidate qualification. • Performed extensive feature engineering (with collaboration customer success managers), negative sampling, distributed hyperparameter optimization (using Parzen Tree and Bayesian Optimization), feature ablation analysis, data annotation / augmentation, etc. • Led development of internal analytics platform to make customer success team more efficient and shed light on key metrics driving the business. Used daily by customer success and executive team.

• Built and trained models for the task of country economic sentiment analysis. • Performed extensive data preprocessing, collection (using mechanical turk service CrowdFlower), modeling, and analysis. Utilized latest Python frameworks (Sklearn, Theano, Scipy, Numpy, etc.), and implemented/benchmarked both shallow and deep models, including Support Vector Machines, Logistic Regressions, Gated Recurrent Unit Recurrent Neural Networks (GRU RNNs), etc.

Worked on textbook rentals team.

Led a section of 8-12 students for the class CS106A: Programming Methodologies. Taught fundamentals of computer science, scored student exams and assignments, and staffed weekly office hours.

Designed and developed a cloud-based note taking and mind mapping software. Applied for a patent on novel ideas.

Startup acquired by Cisco for $260 million in April 2016.