I'm a Software Engineer with a systems-first mindset — I build things that need to be fast, correct, and production-ready, from lock-free C++ engines processing millions of ticks per second to full-stack applications shipped end-to-end.
My core strength is low-latency, multithreaded systems programming in modern C++ (STL, lock-free queues, memory-mapped I/O, CPU affinity), sharpened by years of competitive programming (LeetCode, Codeforces). Alongside that, I actively build in AI/ML and agentic systems — from LLM-driven autonomous software pipelines to hybrid quantum-classical neural networks — because I believe the most interesting engineering problems today sit at the intersection of systems performance and intelligent automation.
I hold an M.S. in Computer and Information Sciences from Adelphi University (GPA 3.84), where I worked as a Graduate Research Assistant on approximation algorithms for large-scale advertising systems, and a B.Tech from NIT Andhra Pradesh.
- 🔭 Systems programming, low-latency engineering, and full-stack product development
- 🤖 AI/ML engineering — LLM integration, agentic pipelines, hybrid quantum-classical ML
- 🧩 Product engineering mindset — I care about correctness, performance, and shipping
- 🏆 Competitive programmer — Codeforces Pupil, active LeetCode grinder
- 📫 Reach me at sunnyrishyvardhan@gmail.com
Roles: SDE / Software Engineer / C++ Engineer
Locations: NYC & NJ area, Hyderabad (India) — open to relocating anywhere in the US
Sponsorship: Requires visa sponsorship (F-1 STEM OPT)
Start: Immediately availableLanguages
Frontend
Backend & Databases
Cloud, DevOps & Tooling
| Domain | Proficiency | Details |
|---|---|---|
| Quantum Machine Learning | Advanced | Built Hybrid Quantum Neural Networks (H-QNN) for medical classification tasks using Qiskit AerSimulator (GPU) and PennyLane; evaluated Lightning-GPU adjoint differentiation for training efficiency; contributor to pennylane-maestro |
| LLM Integration & Agentic Systems | Advanced | Built NightShift — an autonomous software factory using the Claude API that turns a GitHub issue into a deployed, tested PR with zero human involvement; authored custom Claude skills for resume/ATS optimization |
| Approximation Algorithms & Theoretical ML | Advanced | Designed 2-approximation algorithms for the Stream-Ads native advertising problem, validated on a 1.8M-edge YouTube dataset as part of graduate research |
| Applied ML / Data Classification | Intermediate | Model iteration on medical datasets (lung classification, diabetes prediction) with a focus on hybrid classical-quantum pipelines |
⚡ High-Frequency Trading Simulation Engine
A low-latency order-matching and backtesting engine built in modern C++, simulating the core of an HFT system — order lifecycle management, strategy execution, and sub-microsecond matching.
| Stack | C++17, STL (map/list-based order books), CMake, Python (analysis tooling) |
| Scale | Full order lifecycle (new / partial fill / fill / cancel), multiple strategies (SMA, Mean Reversion, Market Making) |
| Performance | Order insertion ~100–500ns · Matching ~200–800ns · Cancellation ~50–200ns — sub-microsecond throughout |
| Security | Integer tick-based pricing (no floating-point drift), RAII/smart-pointer memory safety |
| Impact | Demonstrates production-grade low-latency systems design applicable to quant trading and HFT infrastructure roles |
| Repository | github.com/sunnyrishy/high-frequency-trading-engine |
🧩 C++ Backend Services — Production REST API
A production-style REST API built entirely in C++17 using the Drogon framework, backed by PostgreSQL — showing that C++ can power modern, secure, containerized backend services, not just embedded/systems code.
| Stack | C++17, Drogon, PostgreSQL, libpqxx, JsonCpp, CMake, Docker & Docker Compose |
| Scale | Layered architecture (handlers → repositories → models), connection-pooled database access |
| Performance | Native compiled request handling with connection pooling and structured, timed request logging |
| Security | JWT (HS256) authentication, Bcrypt + salt password hashing, 24h token expiry, protected routes, input validation |
| Impact | Proves out full backend ownership — auth, persistence, and containerized deployment — in a systems language |
| Repository | github.com/sunnyrishy/cpp-backend-services |
🔍 Multithreaded File Indexing & Search Engine
A from-scratch information retrieval engine in C++ implementing an inverted index and TF-IDF ranking to deliver fast, relevance-ranked document search.
| Stack | C++17, STL (unordered_map, custom indexing structures), CMake |
| Scale | Indexes 1,000+ documents with O(N×M) indexing and O(K×D) query complexity |
| Performance | Sub-second indexing, sub-3ms query latency, 10K+ tokens/sec throughput on sample datasets |
| Security | Stateless, read-only indexing pipeline — no external write surface |
| Impact | Demonstrates core information-retrieval fundamentals: tokenization, inverted indices, and TF-IDF relevance ranking |
| Repository | github.com/sunnyrishy/file-search-engine-cpp |
Adelphi University · Nov 2024 – May 2026
Conducted graduate research under Dr. Sixia Chen on approximation algorithms for large-scale native advertising in content feeds, in collaboration with Dr. Sarpe on the StrmAds project — engaging directly with the WWW 2025 paper "Efficient and Practical Approximation Algorithms for Advertising in Content Feeds".
- Designed and analyzed 2-approximation algorithms for the Stream-Ads native advertising problem
- Validated algorithmic performance at scale on a 1.8M-edge YouTube dataset
- Engaged in direct academic correspondence on proof details (Theorem 5) with the paper's authors
Algorithms Approximation Algorithms Graph Theory C++ Python Research
| Role | Organization | Details |
|---|---|---|
| Joint Secretary | E-Yantra Robotics Club, NIT Andhra Pradesh | Led robotics club operations and technical initiatives under E-Yantra, IIT Bombay's national robotics outreach program |
Learning: Distributed systems, quantum-classical hybrid architectures
Building: Agentic AI pipelines with the Claude API, low-latency C++ systems
Exploring: Quantum machine learning, GPU-accelerated circuit simulation
Open To: SDE / C++ Engineer roles — NYC/NJ, Hyderabad, or relocation anywhere in the US
"Build systems that are fast enough to trust and simple enough to explain."