Optimized and maintained data ingestion pipelines: - created cleaning functions for provider (and in-house) data scrapes using pyspark on AWS EMR - optimized merging of many data sources to create a unified dataset (python, SQL) on snowflake - worked with data science team to integrate machine learning models to augment the data (compensation predictions, location mapping, gender/ethnicity predictions, etc.) - built automated quality checks and manually reviewed data before promoting to live (for website backend and other transformation/delivery pipeline inputs)Worked on highly parallelized Go pipelines: - built new features (added customizable inputs, optimized to cut out unnecessary compute, integrate custom appends to pipeline output) - debugged pipeline issues when they ariseDesigned and built data orchestration layer to optimize deliveries on a small team: - Designed and built python library for client facing team to script fully customizable data deliveries - Designed and implemented data orchestration system to automatically read all deliveries and optimize them, so that all data is delivered using as few resources (storage and compute) as possible - Implemented caching so future data deliveries would not have not run in full (or not at all if final result was already available)Company Universe Project: - worked with data science team to design and build Revelio's proprietary company universe - designed and implemented models to merge company identifier data from different company sources to create Revelio Company IDs (RCIDs) - created a company mapper python package to probabilistically map a client's list of companies with identifiers to the RCID universe

• Fixed web app. Debugged HTML, JavaScript (AngularJS), PHP code. Rewrote data calls to google sheets API in PHP.• Reviewed database and made suggestions to further optimize data storage and retrieval.

• Worked on a team of engineers to increase efficiency, reduce variation and ensure high quality of manufactured parts.• Participated in “Problem Solving Meetings” to discover problems in the manufacturing process, identify their root cause and formulate effective action to eliminate the issues.• Designed and ran experiments in the plate shop (zinc-nickel and cadmium electro-plating) to identify sources of variation in production and to improve quality and efficiency. Coded excel work sheets to analyze data. Graphed and presented results to engineering team and production managers.• “Planned Parts”: Read engineering drawings and parts’ lists detailing production requirements. Then, on proprietary software, entered dimensioning and tolerancing specs and planned all processes necessary to manufacture parts.

• Worked with an engineer/project manager overseeing various projects in the WTC complex, including mechanical assembly (e.g., pipe installations & freezes), electrical hardware installations, firestopping, and troubleshooting safety and feasibility issues.• Analyzed architectural, mechanical and electrical drawings to confirm locations, equipment and electronics on the site.• Audited payments to contractors.

Used density functional theory computer models with VASP to assess the viability of doping Cu2ZnSnS4 (CZTS) solar cells with Na and Ca to destabilize antisites in the crystal lattice, thereby, improving efficiency of the cells. Self-taught Quantum Mechanics, with a focus on density functional theory. Submitted manuscript, Berman, S., Gautam, G.S. and Carter, E.A., “Role of Na and Ca isovalent dopants in Cu2ZnSnS4 solar cells” to the journal ACS Sustainable Chemistry and Engineering.

Worked in laboratory of Yair Ein-Eli on a project to optimize lithium battery function. Assembled batteries in a glove box with varying concentrations of perfluorocarbons, assessed oxygen flow, and analyzed and graphed data with Origin.

Designed and built sun-tracking solar panel from scratch