Technology Leader

Built Serverless legal claims processing platform on AWS with Salesforce backend

• •Architected and built an end-to-end serverless legal claims (mass-arbitration) processing platform on AWS with a Salesforce-native backend, automating the entire lifecycle from multi-channel intake through AI-powered proof evaluation to case management and bulk export; delivered as an independent consultant leveraging Claude Code throughout the codebase using Python3.14 – including custom Skills and Hooks, for repeatable, standards-compliance development - in a fast-paced startup environment, collaborating directly with the founding partners, and a business analyst on requirements and prioritization.
• •Designed and orchestrated the core AI proof-evaluation pipeline via AWS Step Functions across document conversion, Textract OCR, and Claude analysis stages, calling Claude Haiku and Claude Sonnet through both AWS Bedrock and the direct Anthropic API, with 7 core Lambda functions and configurable business-rule postprocessing to generate claim verdicts.
• •Engineered 9+ CloudFormation stacks and SAM templates as Infrastructure-as-Code covering storage, database, compute, webhook ingress, and monitoring; built automated deployment/teardown scripts and cross-compiled binary Lambda layers (Pillow, pypdfium2) for arm64.
• •Integrated the platform with SMS Magic, the Jotform API, and the Experian Identity Append API for claim intake and identity enrichment, routed through AWS API Gateway → SQS → Lambda consumers, with JWT Bearer-authenticated Salesforce sync and SSM SecureStrings-based credential management across all third-party APIs.
• •Designed a single-table DynamoDB architecture with GSIs for efficient querying of campaign metadata and claimant data.
• •Built the Salesforce administrative frontend using Lightning Web Components, Apex, and Salesforce Agentforce on a 7-object Master-Detail data model, including bulk CSV export via Lambda/S3 signed URLs and an intelligent contact/claim deduplication service, giving legal and ops teams a self-serve interface for claim and proof tracking.
• •Established CloudWatch alarms and SNS-based monitoring for Lambda errors, DynamoDB throttling, and Salesforce API limit consumption, plus correlation-ID request tracing; authored architecture and developer markdown files to support the small founding team.

Lead AI/ML engineer building Serverless Intelligent Document Processing (IDP) Platform, production LLM systems, and multi-agent pipelines for healthcare operations at scale.

• •Architected end-to-end serverless OCR document processing platform on AWS leveraging Lambda, Step Functions, EventBridge, and Textract to automate enrollment data extraction from healthcare PDFs
• •Designed asynchronous multi-document processing, PDF boundary detection, and parallel execution; validated extracted data against ground truth datasets achieving 96% accuracy.
• •Implemented multi-stage document transformation: PDF ingestion via email (AWS SES) → Textract OCR processing → image conversion → Claude Sonnet 3.7 entity extraction → Pydantic schema-based structuring → CSV template output with validation; integrated IaC (Terraform, CloudFormation) for reproducible drift-free environments across all deployment stages.
• •Automated deployment via GitLab CI/CD pipelines with Docker Containerization and AWS ECR registry integration.
• •Designed event-driven data persistence architecture using S3 for document storage, DynamoDB for metadata and audit logging, and VPC security endpoints for network isolation.
• •Built real-time event streaming pipeline publishing granular events at each transformation stage to Confluent Cloud Kafka topics, synced to Snowflake for analytics visibility and compliance reporting; managed 1000+ daily document operations with production-grade reliability and observability.
• •Implemented RAG framework using hierarchical document chunking strategies, AWS Titan embedding model to generate vector embeddings from chunked segments, and AWS OpenSearch vector database for persistent semantic storage and similarity-based retrieval; enabled intelligent contextual understanding for multi-agent workflows.
• •Evolved platform from single-purpose enrollment processing to multi-agent (document classification, entity extraction & validation), object orchestration system using AWS Bedrock Knowledge Base with RAG semantic search capabilities; extended capabilities to hospital bills, claims processing, and general document workflows with vector-based retrieval.

Built data intelligence and ML systems for global investment operations across 30+ markets.

• •Built full-stack RAG summarization platform combining FastAPI backend (Python 3.12, CrewAI) with 5 specialized pipelines (ODD, Board, Research, PMSR, Putnam Trading) for MosaiQ compute and data platform.
• •Engineered audio-enhanced RAG with AWS Transcribe speaker diarization and Claude 3.5 Sonnet via AWS Bedrock; developed evasion detection classifier using few-shot prompting with 9-category Q&A classification and risk scoring (1-10); persisted outputs in DynamoDB with style personalization and automated citations.
• •Built full-stack Next.js 14/15, Typescript, GraphQL application with Redux Toolkit state management, RTK Query API layer, react-hook-form validation, and Tailwind CSS styling; worked on real-time WebSocket notifications, PDF/DOCX multi-format generation, and DynamoDB-backed ACL authorization with SAML SSO authentication for institutional-grade security.
• •Deployed using AWS provided CI/CD (CodePipeline, CodeBuild, CodeDeploy) orchestrating FastAPI backend, Next.js frontend.

Backend engineer building high-traffic retail systems serving millions of customers. Built 30+ microservices in Java, Kotlin, Python & Go, contributing to Target's digital transformation and cloud migration initiatives.

• •Led the strategic transformation of Inventory & Allocation modules on Target.com from legacy IBM Sterling to a distributed, cloud-native architecture, enabling 5x throughput increase and reducing operational complexity.
• •Engineered high-performance backend microservices using Java/Kotlin (SpringBoot/Micronaut), Go, and Python on GCP, containerized with Docker and orchestrated via Kubernetes to ensure auto scaling and high availability.
• •Led the strategic re-architecture of the Connected Commerce Platform from AWS monoliths to hybrid GCP and event-driven architectures, generating $1B in revenue growth.
• •Built 3+ production microservices in Go using Gin framework with concurrent goroutine-based request handlers, gRPC/Protocol Buffers for inter-service communication; Also engineered integration test data setup pipelines using Golang.
• •Engineered CI/CD automation and infrastructure orchestration: certificate rotation, secret management, and infrastructure provisioning; and managed Jenkins→Drone→Spinnaker→Vela pipeline evolution.
• •Designed distributed systems resilience patterns (circuit breakers, bulkheads, exponential backoff); architected multi-layer caching (Fastly CDN, Redis, Hazelcast) delivering sub-millisecond response times for P1 inventory and guest fulfillment APIs handling 1M+ concurrent member requests.
• •Led Kafka transformation from Zookeeper+brokers to managed Confluent Cloud; designed Avro/Protobuf schema evolution; architected Kafka Streams-based real-time monitoring/alerting pipeline integration with Slack.
• •Implemented reactive programming frameworks (RxJava, Spring WebFlux) and non-blocking I/O patterns across Java/Kotlin microservices; paired with comprehensive observability (Prometheus, Grafana, distributed tracing, logging) and synthetic monitoring (Runscope) for proactive incident detection. Built performance testing framework using JMeter & Gatling.
• •Established enterprise-scale DevSecOps practices: automated vulnerability scanning (OWASP), static/dynamic code analysis (SAST/DAST), certificate management (PEM→JKS rotation), and Vault integration; led security vulnerability management and compliance across open-source dependencies.
• •Led technology integration of acquired companies in stores & supply chain portfolio: established unified coding standards, integrated third-party mobile activations (AT&T, Consumer Cellular, Cricket), and trained engineers on cloud-native patterns.
• •Built reusable libraries for rate limiting, distributed request tracing, and Prometheus/Grafana observability; adopted by 15+ teams with measurable impact (60% reduction in MTTR, 40% reduction in code duplication) enabling self-service deployment patterns across the organization.

CipherCloud was a Cloud Access Security Broker (CASB) and cloud data protection platform, providing enterprises with encryption, tokenization, and DLP capabilities for SaaS applications such as Salesforce, ServiceNow, Box, and SuccessFactors. Led end-to-end development of the product's centralized Management Console — built from scratch — while managing 3 Scrum teams of 12 engineers.

• •Architected and delivered the CipherCloud Management Console from the ground up — a centralized SPA enabling enterprise security administrators to manage encryption policies, data loss prevention (DLP) rules, user activity monitoring, and compliance reporting across all connected cloud applications.
• •Designed cloud application integration layer connecting to Salesforce, ServiceNow, Box, and SuccessFactors APIs; implemented transparent encryption and tokenization of sensitive fields (PII, PHI, PCI data) at rest and in transit without disrupting native application workflows.
• •Built enterprise identity and access management integrations — SAML 2.0 SSO, LDAP/Active Directory federation, and role-based access control (RBAC) — enabling secure, policy-driven access to cloud resources for large organizations.
• •Designed and deployed a responsive single-page application (SPA) using ReactJS, BackboneJS, and Twitter Bootstrap over a Java REST API backend; delivered real-time policy dashboards, audit log viewers, and threat alert workflows.
• •Implemented compliance and audit reporting features, providing security teams with visibility into data access patterns, policy violations, and user behavior analytics across SaaS environments.
• •Led and scaled a high-performing engineering organization through strategic hiring, structured mentoring programs, performance management, and career development planning — growing team velocity and aligning delivery with aggressive product roadmap timelines.

Technical lead for various clients: Union Pacific Rail Roads (UPRR), DEX Media, Hughes Telematics Inc (Verizon Telematics)

• •Technical Stack: Java, Spring, Hibernate, JavaScript, SOAP-based Webservices, JMS, CMS, IBM Mainframe.

Sr. Java Consultant for various clients: GE Healthcare, Lafarge NA

• •Technical Stack: Java, Struts, JSP, Oracle, Jasper Reports, JavaScript.