Open to opportunities

Spurti S M K

@spurtismk

CS undergraduate building backend and ML systems; experienced in Go, distributed systems, computer vision, and GenAI. CVPR-published.

India

Message

What I'm looking for

I'm looking to build high-impact backend, distributed systems, and applied ML products where I can turn research ideas into reliable, measurable production outcomes.

I’m a Computer Science undergraduate at KLE Technological University with hands-on experience across software engineering, backend systems, and applied AI/ML. I currently work on Go microservices, Kafka-based event processing, PostgreSQL, Docker, and Kubernetes at CDPI, alongside research at CEVI in computer vision, 3D reconstruction, and generative AI.

My ML work spans computer vision, medical imaging, LLMs, RAG, and generative models. I co-authored a paper published at the CVPR Workshop NTIRE 2026, where our team ranked 14th among 1,500+ participants. I also ranked AIR 241 among 82,000+ participants in the Amazon ML Challenge and 186 among 2,000+ in the WiDS Datathon.

I’m particularly interested in software engineering, backend/distributed systems, AI/ML engineering, computer vision, GenAI/LLM systems, and research engineering. I’m looking for remote, paid opportunities where I can contribute to real products or research while continuing to grow as an engineer.

Experience

Work history, roles, and key accomplishments

Current

Student Developer

Current

CDPI (eGovernments Foundation & Infosys)

Mar 2026 - Present (5 months)

Migrated 2 Spring Boot microservices to Go within DIGIT 3.0, reducing container memory by 75% and cold-start latency from 15 s to <100 ms. Designed goroutine-based Kafka consumers with bounded channel semaphores, increasing event throughput by >200%.

Go Kafka PostgreSQL Docker Kubernetes Microservices Performance Optimization

Current

Research Intern

Current

CEVI

Oct 2025 - Present (10 months)

Built a SnowflakeNet-based point cloud completion pipeline for 3D reconstruction of degraded heritage monument scans, achieving 0.0142 Chamfer Distance, 78.4% F-score, and 0.74 Normal Consistency. Engineered a multi-stage text-to-3D generative pipeline for heritage and architectural content generation.

PyTorch Stable Diffusion NeRFs 3D Reconstruction Point Cloud Processing Computer Vision SnowflakeNet LiDAR Mesh Generation Motion Synthesis

Full-Stack Developer Intern

Knitspace

Jul 2025 - Sep 2025 (2 months)

Built PostgreSQL full-text search over 21,000+ products using GIN indexes and tsvector/tsquery, reducing p99 search latency from ~800 ms to ~120 ms through query-plan optimization.

NodeJS PostgreSQL REST APIs Full Text Search GIN indexes TSVECTOR tsquery Query Optimization