Open to opportunities

Priyanshu Sharma

@priyanshusharma5

Message

Software Engineer & ML Engineer building end-to-end ML pipelines and LLM-powered applications.

India

Message

What I'm looking for

I’m looking to build production-grade ML/LLM systems with strong engineering practices—clean architecture, fast iteration, and measurable impact on latency, reliability, and model quality—while growing deeper in research-to-production pipelines.

I’m a Software Engineer and ML Engineer with a strong foundation in machine learning, deep learning, and full-stack system design. I enjoy writing clean, modular, well-documented code that solves real-world problems.

In my current role as an Application Developer at Context Mover, I architected a multi-tier context migration system for major LLM interfaces, cutting context-switching friction significantly. I also built a Tier 3 Attention Engine to compute local vector embeddings with sub-100ms semantic retrieval, and I integrated AST-based parsing to improve reliability on low-spec hardware.

Earlier, as a Research Intern at NIT Kurukshetra, I developed deep learning pipelines for underwater image restoration, improving PSNR by 18% and strengthening generalization with SSIM gains. I also build LLM-powered and AutoML projects—using RAG, tool calling, and classical + neural models—to deliver production-minded systems.

Experience

Work history, roles, and key accomplishments

Current

Application Developer

Current

Context Mover

Apr 2026 - Present (3 months)

Architected a multi-tier context migration system to sync session history across major LLM interfaces, reducing context-switching latency and manual copy-paste overhead by 40%. Engineered a Tier 3 Attention Engine using IndexedDB (Dexie.js) and chrome offscreen documents for sub-100ms low-latency semantic retrieval, and integrated tree-sitter AST parsing to improve migration timeouts and failure r

Context Migration Session History Syncing IndexedDB Dexie.Js Offscreen Document API Tree Sitter AST Parsing Low Latency Retrieval

Research Intern

National Institute of Technology, Kurukshetra

Feb 2026 - May 2026 (3 months)

Developed deep learning pipelines (CNN, GAN, and diffusion models) for underwater image restoration, improving PSNR by 18% on benchmark datasets. Optimized custom loss functions and network architectures on paired and unpaired datasets, increasing SSIM by 14% on validation sets.

CNN Gan Diffusion Models Underwater Image Restoration Custom Loss Function PSNR SSIM Paired and Unpaired Training Optimization

Education

Degrees, certifications, and relevant coursework

Bharati Vidyapeeth College of Engineering

Bachelor of Technology (B.Tech), Computer Science Engineering

2023 -

Grade: CGPA: 8.5

Activities and societies: Relevant coursework: Data Structures & Algorithms, Operating Systems, Computer Networks, Database Management Systems, OOP, Software Engineering.

B.Tech in Computer Science Engineering at Bharati Vidyapeeth College of Engineering, Pune (CGPA 8.5), expected to complete by May 2027. Coursework includes core CS and systems topics.

National Institute of Technology, Kurukshetra

Research Internship, Computer Vision / Deep Learning

Activities and societies: Developed underwater image restoration pipelines (CNN, GAN, diffusion); improved PSNR by 18% and SSIM by 14% via custom losses and architecture optimization.

Research internship at NIT Kurukshetra focused on underwater image restoration using deep learning. Built and improved CNN/GAN/diffusion pipelines and optimization strategies to enhance PSNR and SSIM.