# Joshua Kravitz > Head of Technology and Data at United States Senate Committee on Appropriations Location: Stanford, California, United States Profile: https://flows.cv/joshuakravitz ## Work Experience ### Technical Lead @ The Council of State Governments Jan 2024 – Present ### Software Engineer @ Tatta Bio Jan 2024 – Present ### Owner @ Magic Clay LLC Jan 2024 – Present ### Head of Technology and Data, Committee on Appropriations @ United States Senate Jan 2022 – Present ### Technology Policy Fellow @ U.S. House of Representatives Jan 2021 – Jan 2022 | Washington, District of Columbia, United States House Committee on Oversight and Reform, Subcommittee on Government Operations ### Congressional Digital Innovation Scholar @ TechCongress Jan 2021 – Jan 2022 | Washington, District of Columbia, United States Served on the House Committee on Oversight and Reform, Subcommittee on Government Operations. ### Research Assistant @ Stanford University Jan 2019 – Jan 2021 | Baiocchi Lab Under the advisement of Mike Baiocchi (Stanford University) and Jordan Rodu (University of Virginia), we are developing a method to estimate causal effects of behavioral interventions using free text (R). ### CS103 TA @ Stanford University Jan 2019 – Jan 2019 I was a Course TA for CS103, Mathematical Foundations of Computing. I held weekly office hours and graded student assignments. ### CS103 Head TA @ Stanford University Jan 2018 – Jan 2018 Selected among 14 course staff as Head TA for CS103, Mathematical Foundations of Computing in Fall Quarter. Planned and executed course logistics and served as the liaison between the course lecturer and the 13 course staff. ### Data Consultant @ Gavi Begtrup for Mayor Jan 2021 – Jan 2021 ### Deputy Data Director @ Jon Ossoff for U.S. Senate Jan 2020 – Jan 2021 ◦ Supported volunteer (vol) and paid (Community Mobilizer (CM)) relational organizing programs: we built a relational network of more than 160k voters, many of which were low-propensity and first-time voters. ◦ With relevant stakeholders, developed metrics to track performance of volunteer and Community Mobilizer (CM) relational programs: built network of >160k voters, including many first-time and low-propensity voters. ◦ Built dashboard to support management of 3000 CMs by 40 full-time campaign staff, who could then evaluate CMs’ performance and identify key voter networks (SQL, Google Data Studio). Public demo available. ◦ Mobilized 4-person engineering and design team to build OutReach – a phone bank for personal networks – in less than a week, facilitating contact with >17k voters in the final two days of the campaign. Open-sourced. ◦ Published analysis of program on the Analyst Institute: found that program improved turnout by an estimated 3.8 percentage points and was particularly effective on young and low-turnout voters. ### Data Director @ Sri for Congress Jan 2020 – Jan 2020 | Houston, Texas, United States ◦ Supported field team’s data program: pulled phone bank lists in 20+ languages (SQL, VAN), managed technical platforms and their associated data (Reach, OutVote, ThruTalk), and refreshed our volunteer database on a regular basis. ◦ Assisted Finance team in automating repetitive processes, including generating Event and Call Time reports (AppScript), monitoring success of Call Time lists (Google Sheets), and tracking hours of Call Time on the candidate calendar. ◦ Supported and managed team of 5+ technical volunteers and interns, who built out an opponent parody website and multilingual chatbot, cut turf for our canvassing and lit drop efforts, analyzed data from our volunteer Weekends of Action, and more. ◦ Built Early Vote dashboard (SQL, Google Data Studio), which allowed campaign to effectively track vote share and support among target demographic groups in TX-22. ◦ Developed Early Vote data pipeline for our relational program on Reach: an hourly Portal job pulled voters from Phoenix that had voted early but had not been tagged as having voted in Reach and wrote them out to a Google Sheet, then a Google AppScript tagged those voters using the Reach API. This solution made a complex process a simple 5-minute daily task. ◦ Collaborated closely with Texas Democratic Party's data team in ensuring our voter contact targets were utilizing our resources effectively. ◦ Secured over 20 campaign staff – including candidate – and fellows: gave presentation on key security principles, set up staff on password managers (1Password and LastPass) and security keys, and ensured devices were encrypted and up-to-date. ### Political Technology Specialist @ DigiDems Jan 2020 – Jan 2020 DigiDems embeds full-time tech, digital, and data talent on the most competitive races and addresses critical long-term infrastructure gaps with a focus on data and security. I served as the Data Director on Sri Preston Kulkarni's congressional race for TX-22. ### Team Lead @ Bluebonnet Data Jan 2019 – Jan 2020 Team lead for volunteer analytics group working to elect Kelly Stone to the Texas Railroad Commission. ### Co-Chair @ Camp Kesem at Stanford Jan 2018 – Jan 2019 ◦ As one of two student directors, I managed a team of nine that make up Kesem’s executive team and cultivated a volunteer body of over 60 students. ◦ I oversaw the planning and execution of our largest annual fundraiser, Make the Magic, which raised over $28,000 in a single night – exceeding our goal of $25,000. ◦ I worked with our Family Relations Coordinator in recruiting a more diverse set of campers, racially and socio-economically. Through online advertisements and in-person conversations, we ended up with the most racially diverse set of families we’ve had since Camp Kesem’s founding in 2001. ◦ I secured over $1000 in in-kind donations, which included camper water bottles, camp decorations, and a two-way radio rental for leadership communication at camp. ◦ I supported our Programming Coordinators in developing a new camp activity, Part of Me Day, which gives counselors and campers alike the opportunity to learn more about cultures other than their own. We hoped this activity would make camp feel safe and welcoming for campers from all backgrounds. ### Counselor Coordinator @ Camp Kesem at Stanford Jan 2017 – Jan 2018 ◦ As one of two counselor coordinators, I facilitated and managed a volunteer body of over 60 students through community bonding, counselor selection, and counselor training. We worked on several new initiatives in the 2017-2018 school year. ◦ To promote leadership transparency, we wrote monthly counselor newsletters to keep the community informed of leadership’s progress. Counselors reported both reading these newsletters and appreciating the monthly updates. ◦ We anonymized our hiring process so as to reduce bias and promote inclusion. I wrote several software packages in Python to handle the anonymization process, including Python scripts to create anonymous application PDFs and to assign application readers. These practices led to one of the most diverse set of counselors we had had in years. Camp Kesem at Stanford continues to use my work in the selection process. ◦ We moved up the hiring timeline so as to integrate and train new counselors more thoroughly before camp. ◦ We designed additional unit leader and cabin counselor trainings, including ones on QPR, camp scenarios, and difficult conversations. Although I left the position with much work to be done, new counselors reported that these new trainings (in particular, the scenarios training) helped immensely in their preparation for the week of camp. ### Research Assistant @ Stanford University Jan 2018 – Jan 2018 | Sinha Lab Under the direction of Dr. Robert Lerrigo and Dr. Sidhartha Sinha, I analyzed thematic content of online forum for patients with Inflammatory Bowel Disease to better understand patient's emotional states for improved care. We used LDA to qualitatively determine common topics of discussion on the forums. ### Research Assistant @ Icahn School of Medicine at Mount Sinai Jan 2017 – Jan 2017 | Greater New York City Area • Performed data analysis on a cohort of patients diagnosed with advanced urothelial carcinoma, looking for answers as to why patients with low PD-L1 levels respond to anti-PDL1 immunotherapies. • Ran and compared a variety of computational tools which compute immune cell compositions of a given sample. Found that they correlate poorly and don't correspond well with a gold standard. • Attempted to use computational tools to fi nd explanations for low PDL1 patients' responsiveness to anti-PDL1 therapies. • A more detailed summary of this work can be found at https://github.com/jlkravitz/hammerlab17 ### Software Engineer Intern @ Asana Jan 2016 – Jan 2016 | San Francisco Bay Area • Migrated search cluster into Googles Kubernetes framework, making the cluster stabler and more easily scaled. • Developed the override system for a new App Con g framework, giving us a con figuration system more stable to failures and more user friendly (written in Scala). • Worked on small stability-related tasks like killing misbehaving nodes, making staging clusters easier to use for engineers, and making emergency scripts less prone to human error. ### Section Leader @ Stanford University Jan 2015 – Jan 2016 | Stanford University As a section leader for Stanford's introductory computer science courses, I led a weekly section of 10 – 12 students, graded assignments and exams, met with students 1:1, and held weekly office hours. ### Research Assistant @ Stanford University Jan 2015 – Jan 2015 | Stanford Computer Vision Lab • Ran clustering experiment in order to get a better sense of the variety in our data. Conducted on region phrases using mini-batch k-means, found clusters for categories like tennis, water, and transportation. • Developed a pipeline to preprocess and train our data for baseline attribute and relationship classification experiments. Finetuned the 16-layer VGG network using the Caffe deep-learning framework. • Trained and tested the NeuralTalk research project on VisualGenome data. • Read and analyzed computer vision papers. • Co-author, "Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations", IJCV • Co-author, "Embracing Error to Enable Rapid Crowdsourcing", ACM SIGCHI 2016 ### Software Engineer Internship @ CodeHS Jan 2013 – Jan 2015 June-December 2013 (remote work), Summer 2014 (SF office) ◦ Built flexible and extendable autograder which makes it easy to add new tests for specific exercises. ◦ Built new CMS with AngularJS, on which admins could more easily develop and preview course curriculum. ◦ Set up testing infrastructure for front- and back- ends (Django: Mock and Nose, Javascript: Karma and Jasmine). ◦ Helped develop content for professional development course, which focuses on how to teach computer science. ## Education ### Master's degree in Statistics Stanford University ### Bachelor's degree in Computer Science Stanford University ### High School Rio Americano High School ## Contact & Social - LinkedIn: https://linkedin.com/in/jlkravitz --- Source: https://flows.cv/joshuakravitz JSON Resume: https://flows.cv/joshuakravitz/resume.json Last updated: 2026-04-11