Suyash Singh
AI Researcher & Software Engineer.
Bengaluru, IND
About
•
I'm a Computer Science Engineering student and published AI researcher based in Bengaluru, India, building production-grade AI systems, LLM pipelines, and highly scalable backend infrastructure.
•
I focus on memory-augmented agentic workflows, machine unlearning frameworks (GDPR compliance), and high-performance Rust/Go backends.
Core Skills & Stack
LANGUAGES
Rust
Go
Python
TypeScript
C++
SQL
AI & ML SYSTEMS
PyTorch
Machine Unlearning
Hugging Face
LangChain
RAG Pipelines
Embeddings
BACKEND & INFRA
FastAPI
Redis
Kubernetes
Docker
OpenTelemetry
AWS
Prometheus
Experience
Engineered real-time orchestration pipelines and event-driven backend services, enabling low-latency state synchronization and interactive execution for autonomous demo engines.
Optimized asynchronous data ingestion and streaming telemetry across concurrent client sessions, improving backend throughput and eliminating runtime latency bottlenecks.
Engineered distributed, high-throughput asynchronous backend services on AWS, designing state machine transitions, concurrent webhook pipelines, and robust session recovery under heavy traffic.
Implemented resilient routing mechanisms and fail-safe circuit breakers backed by automated testing suites, maintaining high service availability and predictable sub-millisecond execution loops.
Projects
View all
PathFlow
AI-powered production debugging platform that traces execution paths, isolates root causes with evidence, and proposes verified patches.
Semantic LLM Gateway
API proxy acting as an intelligent cache & cost-aware router for LLMs with semantically stored hits (<50ms caching).
Writing & Research
Self-Erasing Neural Networks (SENNs): GDPR-Compliant Unlearning
Published in ICDDS 2025 • IEEE Xplore (Click for details)
Coding Activity
GITHUB CONTRIBUTIONS
AugSepOctNovDecJanFebMarAprMayJunJulAug
LEETCODE
254 Solved
# ask why Suyash matters