Projects
Sandhya has spent over 20 years perfecting her expertise of integrating enterprise software solutions in complex ecosystems with clients from across the world!
​
At Meta, Sandhya resolved at scale analytical challenges by running a large number of Presto clusters spanning data centers across the globe, where Presto is a highly optimized interactive SQL query engine. She designed and led the development of projects to develop cutting edge features for Presto while deploying this highly interactive SQL engine to a “Meta Scale” volume of clusters, ensuring consistent availability and reliability. Sandhya led projects to enhance Apache Spark for batch processing. She was instrumental in bringing efficiencies to MLOps pipelines for gleaning online and offline insights from real-time data streams. This led to revenue increase of over $1 billion.
​
Over the course of 2 decades, Sandhya has designed, developed and deployed Enterprise software in various industries. In addition, she has collaborated closely with defense organizations in the US and UK to develop cutting-edge Tactical Threat Management systems using advanced Mathematics and complex Machine Learning models.
​
At present, Sandhya is using AI/ML to design and develop Risk Evaluation platforms used during Authentication, Authorization and continued Onboarding of SMBs and Large Enterprises, while executing Payment transactions, evaluating Credit Loan applications, Buy Now Pay Later product line, Settling Disputes, Managing Collections at scale and more.
1
Risk Management in FinTech
Developed ML models to provide real time Risk Scores based on behaviors observed during a user’s journey while using PayPal products. The user interactions in the FinTech product flows were captured as Event Series data. The Risk Scores were cumulative and part of the user context, used in real time Risk evaluations. To understand anomalous behavior, along with real time evaluation, a repository of fraudulent ML patterns was used as a continuous feedback loop.
2
Designed and developed Cloud Native Authorization software
that enables the definition, enforcement and management of fine-grained security policies as code across infrastructure, applications and APIs.
3
Enabled Customer Integrations with OpenAM, Okta Single Sign-on and Ping Identity Cloud Platform
Architected solutions to ensure seamless, automated workflows that weave together authentication, user management and Multi Factor Authentication to delight customers without sacrificing security.
4
Handled use cases of seamless signup and onboarding
secure friction free login, account recovery, personalized profile management.
5
Provided automated guardrails
so customers can deploy Kubernetes resources that align with security, compliance and operations team’s requirements.
6
Led the product development of Real Time Gateway software
with teams in US, London and Israel to enforce compliance regulations for Real time screening of Payment Transactions across Facebook platforms namely Instagram, WhatsApp, Oculus, Facebook app, Messenger and more.
7
Developed Risk Profiles for Facebook products
Led the development of software for Fraud Detection, Fraud Monitoring and Dispute Resolution for disputes, claims and chargebacks.
8
Combined Machine Learning models
using Facebook Feature Framework (FFF) for Risk Management, Fraud Detection and Compliance requirements for payment transactions across businesses.
9
Dealt with Exabytes of data flowing across global data centers
Collaborated on the development of Presto (highly interactive SQL engine), Apache Spark (Batch Processing at scale) and Stream Processing Engine with monitoring and alerting built-in for handling peak workloads, thereby serving the Analytics needs of internal and external businesses at Facebook.
10
Enforced Privacy of data
during payment transaction processing with required audits to ensure GDPR and CPRA compliances.
11
Led a matrixed team
to design Architectural patterns for Data Services for consumption by enterprise customers on the Public Cloud including SQL and NoSQL databases.
12
Engaged with Customers to integrate multiple data systems
using diverse protocols and helped them to migrate to IBM’s public cloud. Adhered to various compliances namely FedRAMP, FFIEC, SOC1, SOC2, PCI and HIPAA for Compute, Network & Storage systems in Data Centers worldwide.
13
Led a geographically dispersed team
to Deliver Watson Natural Language Processing APIs on IBM’s Public Cloud.
14
Led a matrixed 200+ member team
responsible for the Design, Development, Testing and Delivery of a HealthCare Cloud Platform, for both GLP (Good Laboratory Practices) and Health Insurance Portability and Accountability Act (HIPAA) compliances
15
Designed and developed the Must Gather Automation
As the Tech Lead for Site Reliability Engineering Team (SRE), designed and developed the Must Gather Automation for Watson Platform for Health GxP along with Real Time Monitoring for the Cloud Platform using Machine Learning algorithms.
16
Led teams to Design and Develop Analytics Tools
for analyzing large data sets to gain new insights for improved patient outcomes, including Spark Architecture for massive scale and in-memory performance.
17
Led a team to Design and Develop a Test Automation System
​for testing the High Availability & Resiliency of all the services in the Watson Platform for Health GxP, while using Kubernetes & running in Containers on the IBM Cloud.
18
Managed relationships across the Watson Health Organization
to migrate to services on IBM Public Cloud namely NoSQL & SQL Databases, Backup, Logging, Monitoring and Alerting. Designed the Resiliency, High Availability and Disaster Recovery for Watson Health applications while running on the Cloud.
19
Led Developer Advocates for evangelizing Watson NLP APIs on IBM Cloud, Bluemix
Completed projects using Watson APIs with Java champions worldwide. Presented at worldwide conferences on the Natural Language Processing APIs offered by IBM Watson.
20
Led the product development of BPM (Business Process Management) software
while leading cross-functional and geographically dispersed teams.
21
Supported Enterprise customers
such as BNYM (a worldwide leader in wealth management), Banco DaVivienda (leading bank in Bogota, Columbia), HSBC, Mexico City, Mexico (one of the largest banking and financial services institutions in the world), and others.
22
Led the development of WebSphere Liberty offering on the Cloud
with a smaller memory/disk footprint and faster startup time as the Lab Advocate for Cerner HealthCare, Inc.
23
Designed and Developed WebSphere 5.0, 6.1 and 7.0 offerings
Led the strategy and product development while integrating the software in complex ecosystems at companies such as Bank of America, Federal Reserve Bank, eBay and more.
24
Initiated a project with the worldwide Java User Groups (JUGs)
This project was shared at IBM's Keynotes at JavaOne, 2016 and at Devoxx, Belgium, 2016. As part of the project, created an AI/ML based Conversational search engine that provides in depth answers to questions pertaining to the Corpus of knowledge on which the engine was trained.various Technologies.
​
The training was done using the videos of presentations from Industry Leaders and publications collected over 20 years of Devoxx Conferences. Leveraged cutting edge Watson Technologies such as Conversation, Speech to Text, Discovery and Natural Language Understanding.