Alex Soloway Email & Phone Number

Newton, Massachusetts, Us

- built RAG document search app using Python (LlamaIndex, Streamlit) and Opensearch to store vector embedding. The app was built to work with multiple LLM providers (OpenAi, AWS Bedrock).- built LLM enabled medical record abstraction pipeline using ChatGPT, AWS Bedrock and Spark NLP Medical.- developed medical imaging app for image segmentation and classification with the segmented features stored in Snowflake and the front end built with Streamlit.

Richland, Wa, Us

- Built custom deep learning network for computer vision based on a modified UNET using Pytorch for image segmentation that improved performance from 80% to 98%. The model has been submitted for copywrite.- Developed data pipeline on AWS and Databricks for daily batch processing of cyber security data from multiple sources utilizings AWS Lambda, DynamoDB, SQS queues and Cloudwatch. Deployed the pipeline using AWS CDK (infrastructure as code) and wrote custom data processing pipeline using Pyspark - Developed an ETL pipeline on Databricks using Pyspark to ingest and process documents from six data sources. Worked closely with data scientists to integrate NLP enrichments (Spark NLP) into the data pipeline and built AWS service to run Mordecai (geoparsing DL model) at scale. Service was deployed on AWS using Docker- Developed python codebase for radar signal processing used to detect oil spills below the ice in the arctic. Introduced engineering best practices like unit testing and git workflow.

San Francisco, Ca, Us

- Designed and built an ETL pipeline to analyze three months (70 GB) of historical bus data (bs4 used for web-scraping) to identify choke points along routes- Utilized Python and PySpark to process the data on AWS using a three node cluster and loaded the results to PostgreSQL for database storage- Produced interactive dashboard to visualize the average travel times for bus routes using Plotly Dash

Seattle, Wa, Us

- Managed project studying noise from bridge expansion joints Successfully identified noise source, saving $6,000,000 by preventing the installation of ineffective noise control measures- Developed a measurement collection and data pipeline to analyze sound recordings using Python (Scipy and Numpy) to build a module for DSP and Fourier analysis, and visualized the data using Matplotlib- Constructed event detection algorithm using scikit and and training dataset to identify and label noise events with an accuracy of 90%, eliminating the need for over 60-hours of manual data labeling- Presented the results of the study and recommendations to a non-technical audience of Washington State Senators and the City of Medina, WA city council resulting in the city pursuing Phase 2 funding

- Analyzed three weeks of sound recordings collected in the arctic to track ice cracking events and earthquake with an uncertainty of %1 using Python and Matlab to develop custom DSP algorithms - Collected and analyzed noise recordings of underwater explosions in Matlab and developed models to predict environmental noise levels with an accuracy of 90% and model to predict seabed properties that showed the Navy’s standard noise prediction model, in use for 20+ years, did not work in shallow water- Authored report used by Navy biologist to determine the environmental impact of noise on marine life and improve environmental guidelines to protect marine mammals from hearing damage- Developed data pipeline to implement an existing computer model in Python and wrote a custom module for post-processing of the data automating the analysis process and saving 5 days of hands-on work per model run- Participated in over 10 at-sea field experiments running data collection systems

- Communicated results of noise and vibration data analysis to non-technical audience- Analyzed multi-day noise recordings using Fourier Analysis to determine noise sources and necessary noise mitigation measures

Boston, Ma, Us

- Working with the modernization division to develop a report outlining the future trends in scope, budget, and schedule of the capital improvement projects for the state-wide housing portfolio.- Performing life cycle cost analysis to determine the value of using high efficiency heating systems.

- Involved in projects designing energy efficient and sustainable HVAC, plumbing, and fire protection systems for industrial construction projects, wastewater treatment facilities and pre existing buildings - Assisted with drafting of mechanical drawings - Ensuring accurate correlation of products chosen by contractor and the related engineering specifications

Doctor Of Philosophy - Phd, Mechanical Engineering

Mechanical Engineering, Underwater Acoustics

B.Eng, Mechanical Engineering