Available for opportunities

Krunal
Jesani

AI / ML Engineer

Generative AI  ·  Computer Vision  ·  LLM Pipelines  ·  NLP

Würzburg, Germany

4+ Years Python
M.Sc. Computer Science
4 AI Projects
Krunal Jesani
Scroll

Building AI systems
that actually work

I'm a Computer Science graduate (M.Sc.) from Julius-Maximilians-Universität Würzburg, specializing in Generative AI and agentic LLM pipelines. My master's thesis involved designing a closed-loop AI workflow where a local LLM autonomously generates, validates, repairs, and benchmarks PyTorch-based neural architectures for image captioning — no manual intervention needed.

I have strong expertise in Computer Vision and NLP, with hands-on experience in Vision-Language Models, synthetic data generation, OCR, and multi-model evaluation. I care about building systems that are reproducible, scalable, and actually deployable — not just demos.

🤖
Agentic AI

Autonomous LLM pipelines with self-repair loops

👁️
Computer Vision

CNNs, image captioning, VQA, OCR

🔬
Research-Oriented

Rigorous evaluation with BLEU, SRCC, PLCC

What I've built

Multilingual Vision-Language Evaluation

Multilingual synthetic graph and image generation pipeline to evaluate visual reasoning in vision-language models across 10+ languages. Benchmarked Idefics and Pix2Struct on multilingual VQA tasks with OCR-based pipelines.

Python OCR VQA Matplotlib Hugging Face
Mar – Jul 2024

Image Quality Assessment

Deep-learning-based image quality assessment workflow using the SQAD dataset. Applied 15+ data augmentation techniques to improve robustness, evaluated with SRCC and PLCC metrics.

Deep Learning PyTorch SQAD SRCC PLCC
May – Oct 2023

Multimodal VR Interaction System

VR-based multimodal interaction system integrating speech, gesture, and pointing inputs for real-time 3D object manipulation in Unity on Oculus Quest 2. Includes probabilistic fusion logic for valid/invalid command handling.

Unity Oculus Quest 2 C# Speech Recognition VR
Oct 2023 – Feb 2024

Technologies & tools

🧠 AI / ML

Generative AI Agentic AI LLM Pipelines Computer Vision NLP Deep Learning Machine Learning OCR Image Captioning Vision-Language Models CNNs LSTMs Transformers

📦 Libraries & Frameworks

PyTorch NumPy Pandas Matplotlib scikit-learn OpenCV EasyOCR Hugging Face Unity

💻 Programming

Python (4+ yrs) C / C++ OOP Clean Code C# (Unity)

🛠️ Tools & Infrastructure

Git Linux Kubernetes VS Code Jupyter Notebook Data Pipelines MySQL PostgreSQL Oracle

Where I've worked

Graduate Engineering Trainee

Alivefoto Incorporation

Jul 2021 – Jan 2022 Rajkot, India
  • Supported software performance optimization and debugging for engineering and image-processing applications.
  • Filed and tracked 100+ upgrade, maintenance, and bug-fix reports to prioritize technical improvements.
  • Developed scalable image-processing and data-processing workflows for automation-oriented tasks.
  • Collaborated with cross-functional teams to test features, document issues, and support AI-assisted workflows.

Academic background

M.Sc. Computer Science

Julius-Maximilians-Universität Würzburg

Mar 2022 – Sep 2025 Würzburg, Germany

Master Thesis: Agentic LLM pipeline for autonomous neural architecture generation, automated validation, and self-repair for image captioning on MS COCO (NN-Dataset / NN-GPT workflow).

Coursework: Computer Vision · Machine Learning · NLP · Advanced Algorithms · Information Retrieval

B.E. Computer Engineering

Marwadi University

May 2017 – Jun 2021 Rajkot, India

Let's connect

I'm currently open to new opportunities in AI/ML engineering. Whether you have a role, a project idea, or just want to talk AI — reach out.

krunaljesani@gmail.com