Projects

Scheduled Kubernetes workload that aggregates user data across databases and APIs, materializes cohort membership in Couchbase, and exposes it through REST APIs for targeted in-app experiences.

Within the MyQ connected-garage ecosystem, I supported and enhanced the Marketing Cohort tool — a daily CRON job on EKS that runs as a long-running workload, queries multiple tables, calls internal and external APIs, updates Couchbase, and feeds REST APIs that drive cohort-based CTA banners in the mobile app.

Automated TAMP account onboarding orchestration — multi-step workflow covering model mapping, SMA/UMA setup, fees, restrictions, and vendor API integration.

Morningstar's Turnkey Asset Management Program required linking new client accounts across custodians, operations, and trading systems. I designed and led delivery of automated onboarding orchestration replacing ~25 hours per day of manual cross-system linking work.

System-to-system API integration automating account-to-user assignment between Morningstar's platform and a third-party financial vendor.

Built .NET Core vendor API integration with event-driven sync, idempotent handlers, error queues, and CloudWatch monitoring — replacing manual cross-system account mapping by operations teams.

ElasticSearch index with DynamoDB Streams pipeline and .NET Core API for near-real-time portfolio and account queries.

Built ElasticSearch index backed by DynamoDB Streams and PostgreSQL sync, exposed via .NET Core API with auto-refresh and stale-data cleanup — powering Blazor WASM UI and third-party platform consumers.

REST scheduling APIs for oil tanker pickup and drop operations — second Kin + Carta engagement after Fifth Third.

Designed and implemented REST APIs and DDD microservices for tanker scheduling at midstream terminal facilities, with Terraform-provisioned AWS infrastructure and OAuth/JWT authentication.

Full product lifecycle on ASP.NET MVC + Web API platform detecting vendor payment errors and supporting audit/recovery workflows for Walmart.

Built ASP.NET MVC web application and .NET Web API for payment audit workflows — SSO integration, Crystal Reports/Kendo UI dashboards, SQL Server data models, and CI/CD from inception through deployment.

Led migration of on-premises investment management applications to AWS with event-driven microservices, Terraform IaC, and 50% faster deployments.

Architected and executed migration of Morningstar investment management workloads from on-premises to AWS — ECS, RDS, Lambda, SQS/SNS, Terraform automation, and phased cutover with a squad of 3–5 engineers.

First Kin + Carta engagement — hybrid migration of commercial banking applications to AWS with DDD microservices, Terraform, and OAuth/JWT auth.

Designed domain-driven microservices in .NET Core, provisioned AWS infrastructure with Terraform (ECS Fargate, ALB, CloudFront), and delivered hybrid migration maintaining business continuity for commercial banking operations.

Full-stack Garuda platform — .NET backend, AWS cloud architecture, PostgreSQL, market-data pipelines, portfolio/research functionality, and production product ownership.

Blanket Finance is my live fintech product for Indian households. Built on the Garuda stack (.NET API + React/Next.js) with AWS infrastructure, market-data pipelines, Razorpay Pro subscriptions, Capacitor mobile apps, and SEO-optimized content delivery at scale. AI-assisted research is one feature among many — not the product definition.

Chakra rating algorithm powering screener, dashboards, and investing ideas across the live Blanket Finance platform.

Designed and implemented proprietary 1–5 Chakra scoring for Indian equities and mutual funds — precomputed rating pipeline with screener integration, contextual explanations, and refresh on earnings and NAV updates.

Separate .NET 8 CLI codebase for US equity data ingestion, signal generation, backtesting, Monte Carlo, and portfolio risk analytics.

Building a headless quantitative research engine in a separate repo — pipeline architecture (ingest → signal → backtest → analyze) with Monte Carlo simulation, transaction cost modeling, and modular strategy plug-ins.

MCP servers, Grafana-integrated triage agents, test/API scaffolding, and locally hosted Hugging Face models for secure enterprise AI adoption.

I lead internal AI platform and developer experience initiatives — production triage workflows using MCP servers and Grafana, test and API scaffolding automation, autonomous backend agents, and air-gapped Hugging Face prototyping with org-wide standards for 15+ engineers.

LLM-powered investing assistant with stock/MF context injection, rate limiting, and multi-turn research sessions at /blanket-ai.

Built context-aware investing Q&A integrated into blanketfinance.com — Groq LLM with platform data context (financials, Chakra ratings, MF details), rate limiting, and navigation links to research pages.