Scale of growth at Sprinklr1. Grew the team from 1 to 160 over 5 years. The team is global with presence in US, Canada, Switzerland, and UK.2. Built > 45 AI first microservices for 5 products (​modern marketing​, ​modern advertising​, ​modern research​, ​modern care​, and ​modern engagement​). These products serve 9 out 10 Fortune 100 clients in 40+ industries globally across 22 languages.3. Implemented large scale, distributed, and highly available analytics infrastructure. Sprinklr AI infrastructure does 5B predictions/day and serves over 750+ AI models with 80+ accuracy.Technology Stack, Achievements, Innovation and Exposure1. Expertise in building state of the art machine learning algorithms for unstructured data. This includes text, images, videos, and voice​. Technologies used - Tensorflow & PyTorch2. Expertise in building highly available, highly distributed, and highly performant AI infrastructure. Technologies used - Kubernetes, Docker, Flask, InfluxDB, Apache Kafka, Apache Spark, and Apache Parquet.3. Scaled AI for Modern Research module to twitter firehose (450 million tweets/day), reddit, facebook, instagram, and over 350 million other news+blog sources. Technologies used - ElasticSearch, Redis and MongoDB.This work has resulted in -1. 3 patent pending technologies2. Consistent “Leadership”​ position in analyst reports - most notably the Forrester Research3. Awarded President’s Club 2020.

1. Investigated commercialization of patent pending technology on autonomous head and eye movement synchronized with real-time speech for social agents.2. Project charter mandated 100 customer interviews covering 10 major industry segments such as Software, Entertainment, Education, Call Centers, Airline, Health Care, Banking, Telecom, Retail, and Food & Beverage.3. Relevant experience include developer, architect, tester, understanding the business model, talking to customers, coming up with a product roadmap, and product positioning.4. Identified product-market fit, determined customer archetypes, constructed channel and payment flow diagrams, and mapped the web customer support ecosystem.5. Developed rapid software prototypes of a low fidelity avatar using WebGL to test the commercial adoption of the technology.6. Presented the results of rapid hypotheses testing to NSF, I-Corps cohort, and I-Corps mentors on a weekly basis. The final Lessons Learned presentation was very well received, and can be accessed at www.vasants.com/I-Corps.html.

1. Worked as the technical lead in teams of three to five on the Survivor Buddy project, and collaborated with colleagues at Stanford University to develop a robot for victim management. Survivor Buddy was cited by Popular Science as a “Best of 2009.”2. The Survivor Buddy robot development included mechanical design, rapid prototypes of physical mock-ups, electronics and control system design, programming with microcontrollers, sensor fusion, communication protocols (Serial Port, I2C, TCP/UDP, and Websockets), and simulation. This work has resulted in two publications.3. Development of system architectures for autonomous social head gaze and proxemic scaling behaviors. This work included implementation of speech, face, and hand recognition/tracking algorithms using machine learning (Decision Trees, Support Vector Machines, and Random Forests), and validation using laboratory experiments. It has resulted in ten publications.4. Design and development of web and desktop user interfaces to control the robot and facilitate communication amongst different stakeholders such as doctors, family members, victim, and structural experts.5. Expertise in research design and implementation of both quantitative and qualitative studies. Examples of methodologies include variable-based experimentation, surveys, field studies, focus groups, eye tracking, in-depth interviews, and logs analysis.