About MeI work as a software engineer at Facebook on the Libra project in Calibra Wallet Infrastrucutre. Before, I was a Computer Science and Economics student at MIT. I am primarily interested in large scale performant systems, financial technologies and distributed consensus.
After graduating from MIT, I was a developer at Cubist Systematic Strategies, Point 72's propietary quant trading wing, working on low latency trade execution.
In the summer of 2018, I worked on Account Security at Instagram, creating a new way to defend IG accounts from hackers.
While at MIT, I did digital currency, network, security and simulation research at the Media Lab.
When I'm not staring at my computer and feeling adventurous, I hop in the nearest Piper Cherokee and take a few friends for a flight around the Bay Area.
I created a Ethereum distributed application (dApp) to allow for decentralized Ethereum derivative trading with a friend at MIT. Specifically, we created a manual order execution exchange for Ethereum Options. Options are trustlessly and securely held in smart contracts on the Ethereum Blockchain, meaning there is no reliance on a third party. We use Oraclize to retrieve provably secure price data, and IPFS to reduce blockchain storage in a cryptographically secure way.
I surreptiously mined Monero on over 1000 remote devices as part of a security experiment. This was achieved by controlling users' CPUs to conduct in browser mining via a Chrome extension.
I ran a market making and taking book for 2 years, when cryptomarkets were more volatile, and had cross-exchange inefficiencies. I initially built a trading bot to find inefficiencies across exchanges using graph theory. This was challenging as assets needed to be distributed and constantly relocated across exchanges at the same time. To serioulsy profit from pricing arbitrage, I needed to do conduct forecasting to place the right assets on the right exchanges.
More recently, I acted as a Marker Marker in several low volume cryptocurrency markets. For a time, I was able to consistently find bid-ask spreads representing up to ~15% of market price. This was magntiudes more profitable than finding pricing inefficiencies.
I built an iOS application which stores your notes in the cloud, but they are all encrypted using private keys from the iOS device secure enclave. Login with touchID or faceID to decrypt your notes, so noone else can see them even if the storage server is hacked.
I worked with Gavin Andresen (Bitcoin's ex lead developer) on the Random Sanity Project at its inception in 2017. We developed a tool to prevent non-random seeding of numbers, which can lead to major security vulnerabilities.
I created the software for a self driving vehicle (with a Hokuyo Lidar) for two autonomous-driving competitions with four friends. The first was to have the vehicle drive as quickly as possible around a ~100m track with no intervention. The second was to have the robot escape an unknown Labyrinth autonomously with no prior knowledge. Specifically, I worked on localization, mapping, safety control, path and motion planning using ROS (Robotic Operating System).
Although design is not my forte, I did create a personalizable website so you can store all your clothes and belogings online. Not only that, but you can try different outfits on a virtual-YOU to see what looks best. Save the outfits that look good.
I built a program to do my Spanish Homework for me. Perdón, Señor Garcia! I had a semesters worth (1000s) of online exercises to do; so I wrote a script to navigate through the exercises and submit all the correct answers for me. I got an A.
I worked at Point72's systematic quant prop trading wing, on a team called Ohms - we provided low latency trade execution for 90% of Point72 and Cubist trades. We are the outbound route to global exchanges and dark pools, primarily handling equities. I spent a long time profiling and reducing latency on Ohms' critical path. I also helped a trading team replay 10 years of market data from multiple exchanges using a custom allocator, lock free Queue primitives, and compression algorithms for back testing.
I worked on the Instagram Account Security Team in the summer of 2018, providing 1 billion IG users with simple and secure access to their accounts. I created a risk-based authentication system which interfaces with tens of billions of IG requests per day. This system uses many indicators to authenticate or revoke trust from accounts.
I orchestrated automated continuous integration tests of the entire Airware cloud with Kubernetes as containerized service. I also decoupled heavy static data from applications and tests, and lead a data transition to distributed networked file storage systems. I also wrote a lot of test cases!
I used a suite of learning algorithms to price products dynamically. Specifically, I optimized product and S&H (Shipping and Handling) costs using a suite of Random Forest algorithms. I designed a max revenue pricing model for online sales by training models using traffic data, product stock and compeititor prices.
I spearheaded a new product feature which scrapes and exploits online hotel bookings’ price inefficiencies. This arbitrating on hotel reservations saved Paribus customers' $50K in its first month. My Hotel Booking saving feature (alongisde Retail and Credit Cards) worked towards Capital One's multi-million dollar acquisition of Paribus. Details Here
I lead a branch of this Y-Combinator graduated company, developing an application for customers to automatically purchase Amazon products over SMS directly. I also automated Zinc's client charges for its automated eCommerce Arbitrage platform
In a team of 4, we built an Android application using RSS (received signal strength) fingerprinting and inferred motion techniques in order to infer indoor localization much better than GPS or wifi can on its own. Used HMMs (hidden markov models), android's accelerometer and receiver to achieve fine grained localization in MIT building 69 better than iphone/android can.
Built a multi-client secure remote file system providing integrity in the face of an untrusted server. This was based on the SUNDR paper, an early blockchain-like implementation.
I worked on enabling Discreet Log Contracts in Lightning, a layer 2 Bitcoin solution. These are forward or future contracts which can be used by investors to hedge against volatile Bitcoin prices. Specifically, I am working on in-channel contracts to allow parties to trade forwards in a scalable manner.
As blockchain systems scale, and network parameters are taken to the limit, network conditions may impact consensus protocols. I analyzed the impact of heterogeneous network conditions on consensus protocols, especially as block time is on the order of seconds. By simulating POW protocols in a heterogeneous peer-to-peer networks, I was able to break consensus by performing an attack with less than 50% malicious hash power. This research aims to raise awareness about disparate network conditions in systems with short interblock times, and suggests alterations to random peering mechanisms such as Bitcoin's.
Semantic infilling involves filling in significant portions of corrupted images. I explored adversarial methods of achieving semantic infilling on faces. I used a DCGAN (deep convolutional generative adversarial network) trained on facial images to reconstruct portions of images.
Our Lab’s goal is to autonomize an on-demand, shared, electric buggy transport service. This is in an effort to replace city bike schemes like Citi Cycles in New York, and revolutionize the 'last-mile' commute. I am simulating travel routes with real, public (subway) and private (Uber) transport data across cities. See an example of our protoyped buggies in Boston: Click Here
I designed a Graphical User Interface to aid pre-emptive maintenance of the cryostatic systems of ALMA’s Front End system antennas.
I conducted research at MIT’s Computer Science and Artificial Inteligence Laboratories (CSAIL) and presented in a scientific symposium at MIT Sloane. The research paper was awarded ‘Highly Commended’ in the UK National Science and Engineering Competition CompSci category.
• Awarded US Private Pilot Licence – flying Piper Warrior PA28-161 mostly in Arizona and Massachusetts
• Executive on MIT Robotics (and Software Lead on Multi-Rover System GUI) - represented team at Gitex 2016, Dubai
• Built Quadcopter Drone with First Person View (FPV) modifications as personal project
• Former Student of the Royal College of Music, Junior Department
• Recorded for Classical Italian Radio, Israeli TV and BBC
• ABRSM Diploma in Violin Performance, and Grade 8 Level with Distinction
• Concert Master and Concerto Soloist of more than ten Orchestras in the UK, Italy, France and the US
• Spanish – Spent 6 months backpacking through Colombia, Peru, Ecuador, Panama and Spain during ‘gap’ year
• German- received highest numerical score on German Nationwide Exam in UK.
• Placed 2nd in 3Red Trading US-wide poker tournament
• Two time Top 10 finish in Susquehanna (SIG) MIT/Harvard tourney
• Invited to train and play with German Semi-Professional Teams FC Quelle and FC Erlangen Brück
• Captain of Phillips Exeter Academy Summer School Team and Seacoast United Captain.
• Member of Varsity Team reaching London Schools’ Cup Final, awarded Football Colours