Skip to content
View tyrobro's full-sized avatar

Block or report tyrobro

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
tyrobro/README.md

Hey, I’m Shaurya πŸ‘‹

while(alive) {
    build(distributed_systems);
    optimize(latency);
    learn(new_things);
    lift(weights);
}

Computer Science sophomore at Birla Institute of Technology and Science, Pilani – K.K. Birla Goa Campus Interested in distributed systems, high-performance infrastructure, networking, and ML systems.

Currently building systems that care about:

  • low latency
  • fault tolerance
  • concurrency
  • decentralization
  • scalability under chaos

I enjoy writing software that sits close to the metal β€” lock-free engines, distributed storage, peer-to-peer systems, and infrastructure that survives bad network conditions and worse design decisions.


βš™οΈ What I’m Working On

🌌 Nyx

A decentralized encrypted messaging engine built in Rust using libp2p.

  • Peer discovery with Kademlia DHT
  • NAT traversal + relay fallback
  • Forward secrecy using HKDF-based key ratcheting
  • ChaCha20-Poly1305 authenticated encryption
  • Fully serverless communication layer

Basically: β€œwhat if Signal had no servers?”


⚑ Hermes

A lock-free high-frequency trading execution engine in C++.

  • ~44ns median tick-to-trade latency
  • 32M+ ticks/sec throughput
  • CPU core pinning
  • Cache-optimized ring buffers
  • Zero hot-path heap allocations

Built because I wanted to understand how absurdly fast systems are engineered.


πŸ›‘ Heimdall

A distributed storage engine focused on snapshot isolation and intelligent routing.

  • MVCC + WAL-based recovery
  • gRPC streaming gateway
  • ML-assisted access prediction
  • Prometheus telemetry
  • Fault-tolerant chunked transfer pipeline

This project started from an obsession with storage consistency and eventually spiraled into distributed systems research.


πŸ“‘ GridMon

Distributed monitoring infrastructure using FastAPI + InfluxDB.

  • Async telemetry ingestion
  • Real-time monitoring pipelines
  • Edge-agent architecture
  • Sub-10ms inference optimization

🧠 Interests

Distributed Systems
High Performance Computing
Computer Networks
ML Systems
Systems Programming
Backend Infrastructure

πŸ›  Tech Stack

Languages

Rust C++17 Go Python

Systems & Infra

gRPC Protobuf libp2p Prometheus Grafana BoltDB

Backend & Data

FastAPI InfluxDB PostgreSQL

Currently Exploring

Kernel-level optimization Consensus systems Storage engines Network protocols


πŸš€ Experience

Summer Intern @ Foundership - The Business Studio

Working on startup-oriented engineering projects and product systems.

Core Member @ Quark BITS Pilani Goa

Managed large-scale event logistics and operations for 500+ participants across technical and gaming events.


πŸ“š Education

πŸŽ“ B.E. Computer Science Birla Institute of Technology and Science, Pilani – K.K. Birla Goa Campus 2024 – 2028

Relevant Coursework:

  • Data Structures & Algorithms
  • Database Management Systems
  • Object-Oriented Programming

πŸ† Certifications

  • Machine Learning Specialization β€” Stanford & DeepLearning.AI
  • Programming with Google Go β€” UC Irvine
  • Game Theory β€” Stanford & UBC

🎯 Beyond Code

You’ll probably find me:

  • lifting in the gym
  • overthinking distributed consensus at 2 AM
  • benchmarking things that absolutely did not need benchmarking
  • reading systems design papers for fun

πŸ“« Connect With Me


fn philosophy() {
    println!("Build systems. Break assumptions. Optimize everything.");
}

Pinned Loading

  1. nyx nyx Public

    A serverless communication system that works through your terminal.

    Rust

  2. Hermes Hermes Public

    Ultra-low-latency HFT execution engine in C++17. Lock-free SPSC pipeline with CPU core pinning, L2-resident order book, and hardware-aware cache-line alignment. 44ns median latency Β· 32.42M ticks/sec.

    C++

  3. Heimdall Heimdall Public

    Distributed storage engine in Go β€” WAL-based ingestion, MVCC snapshot isolation, LRU caching, gRPC streaming, and a Python ML inference tier for heuristic routing. Includes WAL crash recovery, Prom…

    Go

  4. KTlite KTlite Public

    KTlite is a local-first Hybrid GraphRAG system. It fuses semantic vector search with Neo4j knowledge graphs to eliminate hallucinations and enable multi-hop logical reasoning over your documents.

    Python

  5. GridMon GridMon Public

    Distributed hardware monitoring system with cloud-deployed FastAPI backend and Isolation Forest ML for real-time anomaly detection. Optimized compute latency from 200ms to <10ms through async archi…

    Python

  6. Hansa Hansa Public

    Hansa is a lightweight deep-learning model that detects AI-generated images with over 88% accuracy on 32Γ—32 RGB inputs. Built with a compact CNN and optimized for fast inference, it distinguishes r…

    Jupyter Notebook