Depository

Flip Liquide

October 2023 - Ongoing, 6 months

This little www-dot you find yourself on

This Depository is a personal website and blog, built with Next.js, Vercel, Typescript, and a mix of Tailwind and custom CSS. Equal parts CV, Toybox, journal. Please feel free to reach out with feedback or ideas.

Next.jsTypescriptTailwind CSSReactVercelJavascriptHTMLCSSWeb DevelopmentWeb DesignUI/UXSEOContent ManagementBloggingWriting

Senior Protocol Engineer

Govrn

April 2022 - July 2023, 1 years 4 months

Building tooling for digital organizations and their contributors

In my most recent role at Govrn, a platform for reporting and recording contributions to digital organizations, I developed and maintained our backend infrastructure based on Docker, Kubernetes, Datadog, Prisma, Postgres, Redis, and NATS + Jetstream, in addition to ensuring consistency between our smart contracts and our web application. In addition to maintaining and improving our infrastructure, I also grew my areas of contribution to include more of a proudct and feature focus. I was able to quickly propose, design, ship, and operat new features on our platform, including everything from developing feature specs to deploying and running the feature on our infrastructure.

Leveraging Python, Typescript / Node + Express, automated testing, a robust CI/CD pipeline in Github Actions, and quick feedback cycles, I was able to propose, design, and ship key features for our platform like Twitter integration, contribution tipping, and robust contribution visualizations.

For example, I designed and shipped Verification Frameworks, a powerful tool letting different communities configure parameters for verifying member contributions. We anticipated the need for this while researching other blockchain communities. While our users were initially skeptical of the value, its utility became clear once spam contributions started flooding a few users. Contribution verification was able to provide a robust, flexible way for our users to eliminate spam contributions from their communities.

TypescriptPythonNode.jsExpressDockerKubernetesDatadogPrismaPostgresRedisNATSJetstreamSolidityEthereumWeb3Github ActionsCI/CD

Freelance, Senior Protocol Engineer

Valorem

September 2022 - July 2023, 11 months

Building a permissionless platform for writing and settling options contracts on the Ethereum blockchain

Valorem is an options trading platform for writing physically settled options on any pair of ERC20 cryptocurrency tokens. Initially, I came on to help create options pricing and volatility oracles, and to write an on-chain implementation of the Black-Scholes options pricing model.

I eventually was focusing on revamping the core option writing and settlement smart contract. With a focus on ergonomics and gas efficiency, I was able to design and implement a smart contract 10-20x more gas efficient than competing products; this is crucial for applications living on the blockchain, as every time a user interacts with our product, they pay for the gas to run the code. The importance of our users being able to cheaply interact with our protocol is only enhanced by its role as a Decentralized Finance (DeFi) protocol. Leveraging forge/foundry, and testing automation with Github Actions, I drove our test coverage to %100.

I shepherded the core settlement contract through two rounds of a formal security audit, iterating and incorporating feedback as necessary. In order to address feedback from the audit, I implemented a novel probability distribution library in solidity, adapting a novel data structure from computer science academia to ensure fair play in our protocol.

In addition to the smart contract aspects of the project, I also implemented Seaport order validation on the decentralized backend orderbook in Rust. The backend orderbook provides an interface for market makers, takers, and retail buyers and sellers of Valorem options to quote option trades written on the protocol.

SolidityRustEthereumWeb3DockerForgeFoundryCI/CDGithub ActionsSecurity AuditsDeFiData StructuresEthereum Virtual Machine (EVM)AlgorithmsProbability DistributionsGas EfficiencyErgonomicsSeaport

Software Engineer I & II

Microsoft

May 2017 - December 2021, 4 years 8 months

Scaling the Azure Networking control plane at planet scale

After my previous internship on the Azure Software Defined Networking team, I came back for a full time role. My work focused on the scalability, reliability, and modernization of the Azure Software Defined Networking control plane.

Every creation, update, or deletion of virtual networking resources came through our regional controller; each time a customer updated their virtual network, our service needed to drive a corresponding update in our physical network and data plane. My focus during my time in Azure Networking was on delivering scalability and reliability improvements and delivering new SDN features to our product offerings.

I led several efforts to

- remove single points of failure from our architecture, vastly augmenting our reliability

- modernize and transition our services to use an event-driven/PubSub paradigm, enabling new customer scenarios at massive scales previously infeasible

- equip our services for Availability Zone awareness, providing customers with a new level of availability guarantees

- release a new top level resource (Network Profiles) to allow our Azure Container Instance (ACI) customers to access and interact with a rich existing set of SDN features

- mentor junior developers on our team, and establish a strong culture of knowledge sharing and documentation

among others.

C#AzureDistributed SystemsSystem DesignDevOpsInfrastructure as Code (IaC)NetworkingObservabilityEvent Driven ArchitecturePubSubgRPC + ProtobufUnit TestingAutomated TestingReliabilityIntegration TestingScalabilityPowershellWindows.NETAzure DevOps

Intern

Microsoft

May 2016 - August 2016, 4 months

Prototyping scalability improvements in the Azure Networking control plane

Internship on the Azure Software Defined Networking (SDN) team. Every creation, update, or deletion of virtual networking resources came through our regional controller; each time a customer updated their virtual network, our service needed to drive a corresponding update in our physical network and data plane. During my summer in Redmond, I became very familiar with distributed systems concepts like microservice-based architectures, partitioning, load balancing, and gained practical DevOps experience.

My main project was prototyping a new microservice for the regional networking controller. I also got some exposure to the intricacies of operating and debugging planet scale systems while taking a backup on-call shift for our team.

I enjoyed the Seattle summer so much I'd be back the next year full-time.

C#AzureDistributed SystemsNetworkingReliabilityScalabilityWindows

Programmer

University of Michigan

September 2015 - May 2016, 9 months

Programming and designing experiments for a digital ergonomics lab

During my time as a computer science student at the University of Michigan, I worked as a programmer at an ergonomics research lab in the Industrial Operations Engineering department of the engineering college. I aided the research team in implementing experiments which were designed to test various aspects of human-computer interactions (HCI), UI/UX, and usability of mission-critical interfaces.

In order to test the feasibility of using touch screens in commercial airliners, for instance, the FAA commissioned our lab to run experiments on participants in a flight simulator. I integrated an Android tablet on a solenoid-driven mounting arm to simulate the effects of turbulence. Using an Arduino to control the frequency and amplitude of the sinusoid driving the position of the mounting arm and tablet, while the participant was also operating the flight simulator, we would ask them to manipulate and navigate touch screen controls on a custom Android app I wrote and uploaded to the device. This approach enabled us to get good data on the efficacy of touch screens in mission-critical operations.

Another experiment I implemented focused on the integration of an eye-tracking unit into a NASA attention testing application. Study participants would interact with the application, and complete various tasks which would place strain on their attention and focus (e.g. ensuring UI components are correctly configured, answering arithmetic questions, etc). I integrated support for changing the application's appearance in response to participant focus on certain parts of the screen. This involved writing multithreaded code in VB.NET to asynchronously consume data from the eye tracker, and to dynamically alter the UI in response to this. In addition to this, I created a simple "script" format for the researchers to be able to describe different scenarios (e.g. blink this component, flash this one for two seconds).

C#VB.NETAndroidArduinoHardware IntegrationMultithreadingUI/UXHCIUsabilityResearch

Intern

OG Technologies

May 2015 - August 2015, 4 months

Integrating new laser sensor hardware into an automated steel manufacturing defect detection system

OG Technologies is a manufacturing automation company based in Ann Arbor, Michigan. Their products enabled key defect detection scenarios for hot rolled steel manufacturers like ArcelorMittal, Hyundai, and Tata. During my internship, my chief project was to integrate incorporate new Keyence laser imaging hardware into a steel manufacturing defect detection system using C++ STL & Boost, Win API and OpenCV libraries.

This involved learning the hardware sensor's interface for changing imaging modes, reading laser ranging data to and from a memory buffer, and leveraging multithreaded code to return data to the host application in a timely manner. In addition to integrating with our existing software suite, I was able to create a utility & debugging application for configuring the sensor and visually rendering the data being recorded, thus cutting down the time to integrate the sensor in the manufacturing line.

I was also able to implement circle-fitting based defect detection for the data returned from the newly added imaging hardware. Essentially, when imaging rolled steel bars, this would calculate the circle of best fit and then capture the deviations of the profile of the bar from this circle. When integrated with our software suite, this provided our customers with a way to detect defects early in the manufacturing line, thus saving time, material, and money.

C++STLBoostWin APIOpenCVMultithreadingHardware IntegrationAutomationDefect Detection