Building Privacy Vault o Technologies UsedGolang, Java, Python, AWS, CockroachDB, Postgres, ArangoDB, MongoDB, Kafka, ElasticSearch, Kafka Connect, Kubernetes, Istio, Airflow, Spark, S3, MSK, Confluent Kafka, GRPC, PSQL Protocol, Protobuf, lambda, Glue, Rekognition, Okta.o Work◆ Designed and Architected the Vault Solution (PrivacyDB**) - The core data-store which was built from scratch was responsible for storing customer data, honoring Privacy and Security.◆… Show more Building Privacy Vault o Technologies UsedGolang, Java, Python, AWS, CockroachDB, Postgres, ArangoDB, MongoDB, Kafka, ElasticSearch, Kafka Connect, Kubernetes, Istio, Airflow, Spark, S3, MSK, Confluent Kafka, GRPC, PSQL Protocol, Protobuf, lambda, Glue, Rekognition, Okta.o Work◆ Designed and Architected the Vault Solution (PrivacyDB**) - The core data-store which was built from scratch was responsible for storing customer data, honoring Privacy and Security.◆ Designed & Developed the Tokenization product at Skyflow - Supports Deterministic & Non-Deterministic Tokens with UUID, Format Preserving, Length Preserving, and other token formats. - Supports running Analytics on Tokens.◆ Led the Performance Team - were able to do ~25000 tokens per second per customer.◆ Designed an Audit system that monitors and audits every access to Skyflow System(s). Led the team in developing a system that can horizontally scale with 7 years of data retention and DR support.◆ Designed an ETL system to run bulk workloads on the privacy preserved data.◆ Led the API Team - Being an API company, we had to ensure all APIs follow the guidelines.◆ Implemented Initial Authentication (OAuth 2.0) at Skyflow.◆ **PrivacyDB Vault data-store supported - horizontal scalability and provided Disaster Recovery (DR). - Supports PSQL protocol, GRPC, and HTTP to interact with the data. - Fully Encrypted Store (Both InMemory & at Rest) - Supports Polymorphic encryptions by using Symmetric Key, Random AES, Partial Homomorphic and Order Preserving encryption. - Supports Complex real-world objects as a first-class citizen including support for Files. - Supports DLP (Data Leak Prevention) by redacting/masking data based on user's access, via RBAC/PBAC. Additionally Audits each interaction that happens on the data store. - Compliant with PCI Level 1 & 2, SOC 2 Type 1, GDPR, HIPAA, CCPA. Show less

DLP product - document is encrypted and keys at server can be retrieved only with proper access.Implemented ‘Vera for Mail’ from design to customer rollout in under 6 months to win the enterprise license of over USD 3 million with GE and National Geographic.Contributed to integrations with Box, Dropbox along with various other features of the product, some of which related to performance improvements of some core features, Migrated some critical components of a monolithic… Show more DLP product - document is encrypted and keys at server can be retrieved only with proper access.Implemented ‘Vera for Mail’ from design to customer rollout in under 6 months to win the enterprise license of over USD 3 million with GE and National Geographic.Contributed to integrations with Box, Dropbox along with various other features of the product, some of which related to performance improvements of some core features, Migrated some critical components of a monolithic architecture to micro services. Uses Key Value and Search stores like Riak, Elastic Search and Cassandra. Show less

1. hibernate style orm layer on top of no-sql data store 2. custom query language (DSL) to query from the no-sql data storeusing elasticsearch, cassandra, titan and exposing a SQL + hibernate style wrapper so that the platform users can easily create the required models and also execute free form search on them. multi-tenancy in this layer helps app developers to not worry about their data.

design, implement real time chat and presence system

Design and implement features for the product (dimdim).

implemented a product from scratch to manage mobile devices in an enterprise