KEDAR DAVE

I design AI systems that think across text, see, and scale.

Machine Learning Engineer · LLMs · Multimodal AI · MLOps

GitHub LinkedIn Email Resume

My Works

IntelliSearch

Problem

Keyword search fails to capture context in large PDF datasets, making information retrieval inefficient.

Solution

Built a semantic search engine using OCR and vector embeddings to understand document meaning.

Semantic Search MongoDB

Python Flask OCR Vector Embeddings MongoDB

Plottwist

Problem

Manual chart interpretation is slow and lacks metadata.

Solution

Multimodal pipeline using YOLOv5, OCR, and LLMs.

89.2% mAP 70% Faster

Python YOLOv5 OCR LLMs

ImposterCreator

Problem

Deepfake generation required complex tools and expert knowledge.

Solution

Motion transfer framework via First Order Motion Model.

96.4% Success Realistic Motion

PyTorch Jupyter Computer Vision

Sales Analysis Dashboard

Problem

Fragmented sales data across regions hindered strategic decision-making and performance tracking.

Solution

Unified PowerBI dashboard with star schema modeling for 6-month rolling insights.

$163K East Coast Sales 157% Return Rate Analysis

PowerBI Excel Data Modeling DAX

My Journey

Artificial Intelligence Intern

Plutomen Technologies

Sept 2023 – May 2024 · Ahmedabad, India

SaaS Product - Building a RAG pipeline for the internal knowledge base of the company. Optimized vector search retrieval and implemented LLM-based responses.

RAG LLMs Vector DB Python

Machine Learning Engineer Intern

AcmeGrade

May 2022 – Jul 2022 · Bengaluru, India

Detected different types of leaf diseases using CNN classification (ResNet50). Improved model accuracy by 15% through data augmentation and hyperparameter tuning.

CNN Computer Vision TensorFlow

Education

Seneca Polytechnic

Cloud Computing & Admin

2025 · Dean's List · GPA 4.0

Azure AWS Terraform Docker

Humber Polytechnic

Artificial Intelligence with Machine Learning

2024 – 2025 · GPA 3.5

Computer Vision NLP Reinforcement Learning Generative AI

Charotar University of Science and Technology

Bachelors of Information Technology

2020 – 2024 · GPA 3.9

Data Structures Algorithms Database Management

Research

Approaches and Applications of Lane Detection

Analysis of lane detection methodologies for autonomous vehicle progression.

Technical Skills

Generative AI LLM Systems RAG Pipelines Computer Vision NLP MLOps Azure AWS Docker Kubernetes Terraform CI/CD Python C++ Data Structures Algorithms System Design OOP REST APIs Git MongoDB SQL PyTorch PowerBI Tableau Business Analysis