Conducting data analysis, building ML/ statistical models and software data applications using bulk RNA-seq, scRNA-seq and behavioral data to better understand the science of sleep in fruit flys and humans.

Secured pre-seed funding through Desai Accelerator 2022 Program (<5% acceptance rate) in a founding team of 3 and participated in a seven-month immersive experience that involved managing investment, interns, and the startup.Led early-stage startup product development, launching eight zero-to-one product offerings to the App Store in pursuit of product-market fit after securing pre-seed funding in May 2022.Introduced 15 product features, across 8 products, over 2.5 years including a user search (by interest) that used AI to predict user compatibility using Swift and node.js (GCP and MongoDB). Implemented a 'design first' approach to product development cycles, cutting down end-to-end product iteration cycle (ideation, design and build) for an iOS mobile app from 6 to 2 months.Slashed CAC by 80% (from $15.80 to $3.12 per user) allowing us to acquire 5,000+ users by optimizing customer discovery, brand and product roadmaps.Wrote data aggregation (node.js), analytics (Python) and visualization (R) scripts to assess various user engagement metrics that, combined with customer feedback, guided product decisions.

Presented GAP analysis on clients problem resolution processes to ensure Level 2 ASPICE SUP.9 compliance (client was Tier 1 supplier quality management team for a Level 4 autonomous electronic braking system supplied to the largest automotive OEM in the US)Implemented and held workshops on SysML (for internal design & architecture team) using Enterprise ArchitectureReceived trainings in model based systems engineering, requirements management and project management*Was laid-off with 30% of Detroit office due to cost cutting measures

Implemented a logging system for the web service responsible for delivering output 'keywords' for all input 'product skus' to run ads on retailers search page. Designing the system involved interfacing with HDFS, Kafka and Hadoop using Scala, Protobuf and Hive Tables. It is in production to get metrics such as avg. words per sku.Designed a PySpark API that interfaces with FastText, using product information data on the magnitude of 10B, to create vector embeddings that adds depth and context to keywords. FastText is a Facebook AI tool that uses several 'Word2Vec' algorithms to add context and depth to words (i.e. search queries on retail websites) by training on a dNN model. As a result, coverage of keyword embeddings improved from 56% to 98% over the top 50K searched queries.Employed a Scala (Spark) API that takes in as input a location to an annotated feature-set on HDFS and a list of features (including subsets) and calculates the information gain of these features. It is now being used as an internal tool to do basic feature selection for the machine learning model.