Piyush Landge

AI Engineer based in Pune, India. Specializing in Generative AI, Retrieval-Augmented Generation (RAG), and production machine learning applications.

Professional Introduction

I build practical AI applications, LLM pipelines, and intelligent backend systems that solve real problems and run reliably in production. My technical focus centers on Retrieval-Augmented Generation (RAG) pipelines, model evaluations, and high-concurrency Python backend services.

Education and Academic Background

Bachelors of Computer Science

Institution: B.K. Birla College of Arts, Science & Commerce (Affiliated with University of Mumbai)

Period: 2023 – 2026 | Status: Graduated

Core Focus Areas: Data Structures & Algorithms, Database Architecture, Software Engineering, Applied Mathematics.

Technical Skills Taxonomy

Practical Project Experience

Python: Core language, backend scripting, model orchestration.
FastAPI: Asynchronous high-performance REST APIs.
LangChain & RAG: Document chunking, vector indexing, contextual retrieval.
Vector Databases: Semantic search and embedding storage.
PyTorch: Neural network architecture and diagnostic inspection.
SQL & Relational Databases: Schema design, PostgreSQL.
Git & GitHub: Version control, collaborative workflows, CI/CD.

Active Specialization & Learning

Transformer architectures, attention mechanisms, and model fine-tuning sweeps.
Computer vision pipelines and convolutional feature analysis.
Agentic workflows, tool-calling systems, and reasoning loops.
Docker containerization and cloud AI infrastructure.

Featured Engineering Projects

AI Flyer Designer

Category: Generative AI & Design Tool

Problem: Creating marketing flyers manually requires steep design tool learning curves and tedious formatting.

Implementation: Built with an asynchronous FastAPI backend for layout synthesis paired with HTML5 Canvas for real-time visual drag-and-drop layer adjustments.

Verified Results: Sub-second template synthesis, zero-latency canvas layer manipulation, and responsive rendering.

Live Demo | GitHub Repository

Contextual RAG Assistant

Category: Generative AI & RAG Projects

Problem: Standard language models hallucinate on private enterprise documentation and cannot cite verified source evidence.

Implementation: Built end-to-end semantic retrieval with LangChain, PyTorch embeddings, high-dimensional vector databases, and cross-encoder reranking.

Verified Results: Zero-hallucination grounded retrieval on custom documentation sets and sub-250ms vector query latency.

GitHub Profile & Code

Deep Learning Auditor

Category: Machine Learning Projects

Problem: Neural networks silently fail during training due to contaminated data, vanishing gradients, or activation saturation.

Implementation: Python diagnostic modules using PyTorch and NumPy that inspect activation tensors, backprop gradient norms, and partition boundaries.

Verified Results: Automated pre-training data leakage detection, gradient anomaly alerts, and 30% reduction in debugging cycle times.

GitHub Profile & Code

Verified Professional Links

GitHub: https://github.com/piyushlandge04
LinkedIn: https://www.linkedin.com/in/piyushlandge04/
Portfolio Home: https://www.piyushlandge.site/
Resume (PDF): https://www.piyushlandge.site/resume.pdf

Contact

Email: work.piyushlandge@gmail.com

Location: Pune, Maharashtra, India