Distributed Bitcoin Miner

• Building a distributed server-client communication bitcoin mining system

• Programming a custom Protocol built on UDP in Go to provide reliable communication and traffic control between echo servers/clients

• Implementing client-server model with Go-style concurrency and marshalling

• Handling failures, layering, and protocol multiplexing to ensure scalability and consistency

• Github link: DistributedBitcoinMiner

• Implemented Raft, a more understandable alternative consensus algorithm to Paxos , in Go

• Defined routine for leader, candidate and follower, implemented leader election, log replication to provide fault tolerance, data replication and consistency

• Used only Go Channels to achieve asynchronous consistency instead of Mutex

• Github link: Raft_Implementation

• Collaboratively developed a large-scale system in Go with multiple different components communicating via RPCs, allowing users to make posts, subscribe to users and see others’ posts

• Implemented a general backend key-value store based on consistent hashing and sharding to enable request routing and fault tolerance

• Implemented a caching protocol with Lease Callback mechanism to improve throughput while keeping cached data consistent

• Github link: Tribbler_Implementation

CMU Computer Science Academy

• Worked in a team to create a CS education platform for high school students in Pennsylvania, now adapted by 14 high schools

• Built interactive animation/interface for curricular content in Javascript and Django

• Involved in website creation which launched a pilot version used by 400 students, collected user feedback to improve the system

• Github: https://github.com/cmu-cs-academy

• Website: https://academy.cs.cmu.edu/

Natural Language Processing Q&A algorithm

• Built a Python Question-Answering System by scraping Wikipedia content, generating features with the Stanford NLTK package, parse tree and incorporating the trigram model

• Utilized POS tagging, tokenization and various NLP techniques to parse and generate questions/answers

• Generated questions and answers with highest relevancy and won best Q&A algorithm among 20 teams against other grad students