Ashok Gupta Email & Phone Number

San Jose, Ca, Us

San Jose, Ca, Us

More TBD

San Jose, Ca, Us

▪ Leading globally distributed teams to build the Intent & Natural Language Understanding Platform with capabilities for Language Detection, Spellchecker, Cross-lingual Concept Dictionary, Creative Knowledge Graph, Tokenization, Stemming and Lemmatization, Dependency Parsing, Color, Background, Asset Type and Design Type NER, Query Suggestions, Topical Ordering, and Intent understanding▪ Leading multiple work-streams involved in solving search and discovery problems such as SGE(Search Generative Experience) and Intent to Action using cutting edge SOTA technologies in ML(Machine Learning), NLP(Natural Language Processing), AI (Artificial Intelligence) Assistants with LLMs(Large Language Models) and RAG(Retrieval Augmented Generation).▪ Led roll out of semantic search for Discover Panel on Adobe Photoshop and Adobe Illustrator.▪ Led Quick actions search and recommendation.▪ Led App cards and app quick links projects for Creative Cloud Desktop and helpx.adobe.com▪ Spearheading multiple search and discovery initiatives for Creative Cloud Desktop and Creative Cloud Home.

San Jose, Ca, Us

▪ Leading Search and Recommendation APIs for Creative Cloud Experiences.▪ Launched Search and recommendation API for Illustrator Discover Panel for Adobe MAX 2021.

San Jose, Ca, Us

San Jose, Ca, Us

▪ Led and launched Photoshop Discover Panel search and recommendation APIs for Adobe MAX 2020. ▪ Overall project management and technical leadership, collaboration with stakeholders across multiple orgs. Ownership of end to end API success.▪ Critical and significant hands-on contribution: Content ingestion for index, api design and implementation, search and recommendation results evaluation.▪ Operational, development, design excellence with no production issues from start to finish while reducing technical debt significantly ▪ Significant and visible ramp up of Junior engineers through correct processes, documentation and code reviews with successful remote collaboration as focus.▪ Building and driving best practices and positioning the team for accelerated success for wider and deeper future initiatives.

Seattle, Wa, Us

▪ Designed and built additional apis to increase content engagement.▪ Extended backend with launch of additional content providers and various device types.▪ Launched in additional locales and countries.▪ Designed database schema for scale(~20k requests per second)

Seattle, Wa, Us

▪ As the founding member and the technical lead of the project, bootstrapped and launched server side of Alexa multi-model ambient home experience. Supported and launched numerous Alexa device types with a screen.▪ Came up with APIs specifications, built system design and architecture.▪ Directly worked with leadership (Director, Product and Program managers across multiple orgs) and various stakeholder teams such as content providers, devices' and platforms teams.▪ Built critical portions of the server side stack, took care of operational excellence, security certifications and privacy compliance.▪ Guided and mentored junior engineers with diligent code reviews, alignment of their individual tasks with overall design

San Francisco, California, Us

▪ Singlehandedly designed, developed and launched a model-agnostic, real time personalization framework for freelancer profile search and recommendations.▪ The framework facilitates code-less model deployment making it possible for data scientists to do rapid model iterations. Allows deploying multiple models and evaluating them side by side.▪ Did several rounds of performance benchmarking and enhancements e.g. GC tuning, Solr Cache optimizations and Redis data tweaking to make sure the solution serves the real time use cases without any noticeable increase in the latency of the API.▪ Wrote bullet proof, modular, easily maintainable, well documented, fully auto tested code▪ Working on open sourcing some portions of the framework.

San Francisco, California, Us

▪ Created and maintained a set of Micro Services to support indexing and searching of messages, rooms and files, for a real time messaging platform. ▪ Leveraged ElasticSearch as the backend, allowing a 1s refresh(time to search) over a terabyte of indexed data containing over 200 million searchable documents.▪ Tech Stack : DropWizard, Hystrix, Archaius, Jersey, ELK(Elasticsearch, Kibana, Logstash), Graphite and Grafana.▪ Singlehandedly setup and managed Elastic Search cluster using Chef and AWS Opsworks including backup, recovery and security of the indexed data.▪ Performed benchmarking, optimization, implemented computing the MRR of the search results▪ Built a dialog segmentation subsystem using Cascading and Oozie▪ Implemented a simple TF-IDF based endpoint for extracting the keywords from a multi-party chat room

San Jose, California, Us

Low Latency Aggregation: LoLA is a set of Storm Trident topologies to compute near real time aggregates on streaming data. OLAP servers generated reports from these aggregates.▪ Built components of the Low Latency Aggregation Platform.▪ Created storm trident topologies in Scala to compute hierarchical multidimensional feed for OLAP servers from the streaming events.▪ Used Kafka for consuming the data, Cassandra to back the intermediate aggregation state and Zookeeper for the cluster coordination.▪ Created a library to test and verify aggregates using MS Excel pivot tables. ▪ Contributed substantially to define the Event Data Model for the Data Platform.

San Jose, California, Us

▪ Worked first on DSP(Data Services Platform) and then created BDP (Batch Data Platform) for generating reports and ADM(Analytical Data Model) views.▪ Responsible for the overall technical leadership, technology evaluation, design and optimization of the BDP.▪ Fine tuned Hadoop, Hive, Oozie and Sqoop configuration for optimal performance. ▪ Designed tools around the frameworks to make it easy to modify, deploy, run and track jobs. ▪ Implemented monitoring and alerting for the cluster.▪ Refined and implemented an "Universal Context" creation algorithm using Cascalog (a clojure library for cascading) to build the graph closure of an user's data.▪ Design and implemented Jobs to generate much valuable ADM views which could be customized for diverse clients.

San Jose, California, Us

▪ Built a Reports Visualization tool from scratch using Dojo and OSGI and integrated it with in house chat platform.▪ Built DSP (Data Services Platform). Created a fault tolerant, scalable and customizable DSP to generate business critical Reports and ADM(Analytical Data Model) views processing millions of events daily for multiple clients.

Redwood City, Ca, Us

Redwood City, Ca, Us

▪ Turned around Gatherer, Yodlee's data aggregation platform. ▪ Migrated Gatherer from MS JVM to Sun JVM resulting in a great reduction in operational costs.▪ Scaled to 6 million refreshes per day. ▪ Supported new data models for new agents▪ Other projects: Agent Deployment Tool, MoneyCenter, Universal Biller Directory, Payment Monitor, Payment Engine.▪ UBD: Implemented OFAC check for Billers. ▪ Paymon: Built a tool to generate and execute SQL through an UI. ▪ Payment Engine: Worked on the PayPal API integration with Yodlee Payment Product.

Austin, Texas, Us

▪ Completed design, implementation, unit testing and integration of some complex modules such bonds and links for Oracle Beehive.

Master Of Science - Ms, Computer Science (Specialization - Machine Learning)

Bachelore Of Technology, Computer Scince And Engineering

Higher Secondary

High School