Mar 2023 – Nov 2024 · 1 yr 8 mos · Remote

• Achieved FIPS compliance for document verification services by implementing secure cryptographic practices, migrating the microservices from Elastic Beanstalk to EKS (Kubernetes), and provisioning AWS resources with Terraform. These efforts were instrumental in securing a public sector deal by delivering a robust and compliant solution.
• Enhanced the Webhook Notification processor by enriching the product logic, decoupling document verification to significantly improve latency performance and ensure seamless real-time notification delivery, even when client endpoints encounter unavailability.
• Regularly reviewing metrics and addressing technical debt is essential for preventing potential incidents and maintaining product stability. By proactively mitigating risks early on, we ensure our infrastructure remains reliable and scalable.
• Integrating Berbix reduced document verification response time from 16 to 4 seconds. Pre-verifying documents further reduced the time to under 1 second, minimizing resubmit calls from CaptureApp. This has resulted in faster and more efficient document processing.
• Implemented a data deletion script to remove document images and PII older than 3 years, including historical military IDs, ensuring compliance with state regulations. This initiative protected our company from legal and financial risks.

Jun 2020 – Mar 2023 · 2 yrs 9 mos · Remote

• Decoupled the DocV Orchestra microservice to independently manage features, configurations, and logic, separate from the main public-facing service, enabling greater flexibility in adding new vendors and maintaining full control over all document verification processes.
• Implemented BYOK (Bring Your Own Key), allowing encryption keys to be securely encrypted and decrypted using customer-provided KMS keys. This ensures that customers maintain full control over their data, with Socure providing the necessary infrastructure to safeguard it.
• Implemented metrics standardization across all DocV services and integrated them with Datadog for creating dashboards to analyze and monitor metrics. Configured alerts to maintain SLA.
• Built an analytics engine to proactively analyze document metrics collected from CaptureApp, aiding in understanding customer experience issues and resolving them efficiently.
• Addressed vulnerabilities in all DocV service source code by leveraging WhiteSource and AquaWave tools, ensuring compliance with GSA requirements.
• Implemented a configurable batch job for algorithm evaluation as per the requirements of the DS team.
• Developed a customization tool for DocV CaptureApp, enabling clients to create a unique capture app
• tailored to their needs for uploading and verifying documents during customer onboarding.

Jun 2014 – Jun 2020 · 6 yrs · Chennai Area, India · On-site

• Developed an image-based email search with ML text extraction, caching text in Redis against the hashed image to prevent redundant processing and ensure faster access.
• Implemented Missed and Wrong contact suggestions during email composition using a statistical approach, with contact data and statistical models managed in HBase.
• Integrated language detection, sentiment analysis, mail classification, and entity extraction models into the application, enabling features like event reminders, language translation, and email spam processing.
• Developed and implemented a robust library collection for parsing MIME, DOM, and Date content, utilizing context-free grammar.
• Developed a regex validation tool to prevent backtracking and implemented Antisamy XSS filtering for email content, significantly enhancing security.
• Asynchronously detect and highlight entities in email content using a decoupled microservice for seamless integration with other applications.
• Boosted Server performance and maintained backward compatibility by utilizing Google's Protocol Buffers in the cache layer.
• Enhanced code maintainability, scalability, and testability by implementing the layered architecture pattern.
• Maintain the resilience and integrity of email processing by implementing the Observer Pattern with a retry mechanism across all modules.
• Implemented security annotations in the business layer services of all email modules to validate the user
• scope for accessing email content and ensuring proper access control.

Dec 2013 – Apr 2014 · 4 mos · Chennai Area, India · On-site

• Developed a validation tool for SQL queries used in Zoho, utilizing the Context-Free Grammar approach. Implemented DFA (Deterministic Finite Automaton) implementation to eliminate backtracking for regular
• expressions used in the Zoho mail services.