Pushkar Joshi Email & Phone Number

California, United States

Mountain View, California, Us

Dec. 2023 - present: Generative AI for industrial roboticsInitiated large research project for using foundation models to generate all aspects of robot solutions (robot process, skill module code, documentation, digital workcells). Details to come soon.-----------------------------------------------------------Sept. 2019 - Dec. 2023: User-facing tools for industrial roboticsTechnical lead and manager, UX engineering team (user-facing frontends + SDKs) and Technical lead for robotics workcell authoring, representation, and simulation.Responsible for the Intrinsic frontend and SDK tools, i.e. the primary user experience for the Intrinsic platform. Launched externally and deployed at partner sites as "Flowstate" (https://intrinsic.ai/flowstate/). My deliverables include data portals, 3D workcell design, equipment onboarding and configuration, robot application authoring, and robot capability (skill) authoring, all in a single coherent web-app powered by the Cloud. Building a unified experience required UX engineering to interact deeply with the entire Intrinsic technical stack (e.g. Kubernetes, Motion Planning, Perception, Physics Simulation) for industrial robotics. I hired, grew and manage the UX engineering team. My day-to-day involves ensuring progress of technically deep software projects, writing design documents, developing production-grade software, creating product prototypes, and negotiating features and timelines with stakeholders across Intrinsic.

Mountain View, Ca, Us

(worked on team that became Intrinsic LLC)

Mountain View, Ca, Us

Staff Software Engineer: TFX (www.tensorflow.org/tfx)Oct. 2015 - Sept. 2019Founding team member, Tech. Lead for Frontend and Experimentation Infrastructure. TFX is an end-to-end platform for large-scale machine learning, heavily used in production at Google. My work required deeply understanding and optimizing ML infrastructure at Google. I initiated and led two projects: (1) Frontend and (2) Experimentation Infrastructure. (1) The TFX Frontend is the primary interface to TFX. It is a unified and guided interface to all phases of machine learning and provides novel visualization tools for data, model-quality, and pipeline status. I was responsible for all aspects of the frontend, from designing workflows, developing and reviewing code, and fixing breakages. (extended abstract + poster presented at SysML 2018: https://mlsys.org/Conferences/2018/index.html#posters)(2) The TFX Experimentation Infrastructure built tools for experimentation and modeling in TFX. Most tools were for internal Google engineers to change their ML pipeline (config or code), re-train, and observe and track the effect of their change. -------------------------------------------------------------------------------------------Senior Software Engineer: Google Web Designer (www.google.com/designer). Sep. 2012 - Oct. 2015Founding team memberLed sub-teams to deliver projects like:Events (event handler authoring so users could make richer, interactive documents)Templates (fully functional starter documents on multiple platforms and screen sizes)Output File Optimization (cleaning up and optimizing CSS and 3D transforms)Drawing Tools (pen tool, canvas runtimes, etc.)

Chicago, Illinois, Us

Research and development of advanced graphics technology for simplifying graphic design or creating visually rich user experiencesDeveloped the Pen tool (vector design) and Brush tool (raster design) for Motorola Ninja: a graphic design web app. for creating HTML5 web content.

San Jose, Ca, Us

Research and development of innovative features for Adobe products, with focus on curve-based 3D shape modeling to simplify 3D modeling for novices. Wrote the 2D & 3D geometry engine for the 'Repousse' feature in Photoshop CS5 (now renamed to 3D Extrude). Also, created research prototypes for: multitouch user interaction, intuitive shape deformation, and reverse-engineering shapes into easily editable representations. Published several papers and patents.

Emeryville, California, Us

Research in character rigging: implemented extensions of a new articulation technique and wrote the code to incorporate it into the existing rigging pipeline.

Research for realistic facial animation and keyframe modeling

Part of the founding team for an online magazine about the role of engineering in everyday life

Redmond, Washington, Us

Designed and implemented back-end test software for the MSN Money website

Ph.D., Computer Science (Graphics)

Bs, Computer Engineering And Computer Science