Recursion is a clinical-stage biotechnology company with the mission to decode biology to radically improve lives. By integrating technological innovations across biology, chemistry, automation, data science and engineering, Recursion aims to industrialize drug discovery.By leveraging new technology, Recursion creates virtuous cycles of learning around datasets to build the next-generation biopharmaceutical company. It’s complex biology, decoded.

Ezra (www.ezra.com) is a company dedicated to providing early cancer detection for everyone by using the most advanced medical imaging technology. Its ultimate goal is to offer a whole-body MRI scan in a single session and analyse it using the Ezra AI to quickly and accurately screen for precursors to cancer, cancer, and major disease states.• Managed a team of 5 (3 full-stack, 1 MLOps, 1 QA tester). Held weekly 1:1 meetings, conducted performance reviews, interviewed, hired, retained and mentored talent. Led user interviews, collaborated with and communicated technical details with non-technical stakeholders• Architected, built, and deployed a machine-learning solution that converts sophisticated radiology reports to a layman's summary (see BusinessWire article), saving medical staff 75 minutes per report, a 600% productivity boost and a 72.7% cost reduction• Implemented model and API testing suites and added them to our CI/CD pipeline (using DVC, S3, and GitHub Actions), allowing us to build and release with high velocity• Monitored deployed models using CloudWatch, SNS, Sentry, and Slack API• Built and deployed retraining pipelines allowing us to continue to improve the medical staff's efficiency• Built Python RESTful APIs for receiving radiology reports and initiating the ML inference pipeline• Led the engineering team building the non-machine learning components of the product by setting priorities, reviewing code, conducting design meetings, planning release cycles, and sprints• Took the initiative of creating and writing a company tech blog post with an average of 5K views per article, boosting Ezra’s SEO and aiding with recruitment• Member of the Ezra culture committee, helped with organizing in person and virtual events to build organization-wide camaraderie

Aggregate Intellect or A.I. Socratic Circles (AISC) is a community of intellectually curious individuals centred around technical review and discussion of advances in machine learning.As part of the Machine learning in healthcare discussion group lead, I organized presentations, invited guest speakers and presented recent papers in the field.

Ideal empowers talent acquisition teams to make data-backed, high-volume hiring decisions using artificial intelligence.• Designed, implemented and deployed machine learning models that allow Ideal’s screening engine to detect bias in the model and explain its output decisions using Keras, Pytorch, Sklearn• Built a Python RESTful API that recommends suitable candidates for newly created job postings and exposed it through a Node.js proxy

I conducted my master's thesis study at Bloorview Research Institute which is affiliated with the University of Toronto. My research focused on Brain Computer Interfaces (BCI), which involves designing communication systems that are controlled with the brain. These systems can be used by individuals with communication challenges, e.g with cerebral palsy, brain stroke, etc. Using machine learning and natural language processing, I developed a faster and more robust virtual speller controlled by the user's brain signals as my master's thesis. After gathering and analyzing data on 10 participants, this system gained an average 42.84% increase in typing speed and a 4.66% improvement in precision.

Prepared and presented lectures for tutorials of +250 students on normalizing, tagging, feature extraction and classification of political comments on reddit forums using python (spacy, nltk, scikit learn).Developed python scripts for auto-grading assignments.

Prepared and presented lectures for tutorials of +30 students on a variety of topics including regression, classification, clustering, gradient descent, neural nets, tensorflow, numpy.Provided guidance to students about lecture material and assignments.In each assignment students implemented and compared algorithms discussed in lectures on different datasets using python and tensorflow.Prepared and marked assignments and exams

Prepared and presented lectures for tutorials of +30 students on a variety of topics including regression, classification, clustering, gradient descent, neural nets, tensorflow, numpy.Marked assignments and exams.

Designed and marked projects and assignments which focused on business analysis; including customer segmentation based on likelihood of accepting a bank loan, predicting successful or unsuccessful sales and auctions, income rates, etc.