Alok Rao

A mobile app for playing group games—right inside a chat.

On its surface, Blend is a fully-featured chat app. You can create groups and have conversations—the whole shebang. But what you can also do is select a game, send it as a message, and see your chat screen turn into a game board!

In Blend, each action you take in a game, whether playing a card or casting a vote, is transmitted as a message. It travels as a protobuf packet, using a home-grown binary-wire protocol implemented on top of TCP. On the server is a chat engine built to route messages to members, utterly agnostic of their contents.

In the app, games are implemented on an extensible turn-based state machine with re-usable actions. It currently has five completely original games of multiple genres and mechanics, with two more in the works.

The app is built on Flutter, and the chat engine and API are coded in Node.js, NestJS, MongoDB, and Redis.

A blockchain platform for issuing, transferring, and trading financial instruments such as ETFs and private equity.

With two flavors, one built on Hyperledger Fabric, and the other on R3 Corda, the platform runs a permissioned, private blockchain network across organizations of interest. The blockchain has smart contracts written in Go/Kotlin for administration, on-boarding, matching, and similar functions.

The platform has a privacy layer built on top of the implementation to obfuscate transactions and anonymize participants.

The web app is built on React and Node.js, and the platform runs on Docker containers in a Kubernetes cluster, communicating through a P2P network that uses GRPC/AMQP.

A composition framework for web services.

Written in Spring Boot, this framework provides a DSL for rapidly wiring up complex flows with several integration points. The primary abstraction of the framework is a "Task Chain." The implementor uses the DSL to declaratively define the flow of the service. Each task in the chain is a bean that implements a "task processor." Task processors can be extended from off-the-shelf components that simplify calling an API, connecting to a database, or an MQ.

The framework was internally implemented across teams and shaved over 50% of the development time of several APIs.

A fully functional prototype on the Apple TV for financial advisors and clients.

Built on tvOS, the application is a productivity hub for Advisors with several spokes. One is a conversation screen that aggregates your communications with clients across providers such as Email, Facebook, and Twitter. Another is a news source with intelligent summarizations and insights on impact to portfolios. It also has other features such as customer insights, portfolio, and monitoring.

The app was built with Swift and fetched data through REST APIs coded with Node.js connected to a PostgreSQL instance.

A utility service to store, throttle, and guarantee delivery of transactions.

Backed by an ActiveMQ cluster, APF guaranteed high availability through automatic failover and instantaneous message replication. It was a multi-tenant platform that could be configured to transmit transactions in parallel or sequence and formed the bridge for several interactions between mainframe apps and distributed services.

The service was written in Java using Spring Integration and Spring Boot and used ActiveMQ and later Kafka for persisting transactions.

A portfolio tracker for historic analysis.

The app lets you create portfolios from existing financial instruments and track their value across time ranges. Historic data was run through financial models to predict the success of portfolios and tweak their composition.

The app was built in AngularJS with a REST API in Spring MVC.

A suite of Java and .NET web services for processing account opening and transfer of asset transactions.

At the core of the account opening pipeline, the services orchestrated several utilities such as eSignature collection, rule engine validation, faxing and mailing integrations, and manual workflow interactions.

The suite was a collection of services ranging from top-of-the-line ASP.NET core and Spring Boot REST APIs to legacy services running Spring MVC with XML configurations using SOAP.