Machine Learning Engineer with ML/NLP Focus

SHIVAM ARDESHNA

+1-438-***-**** — Montreal, QC, Canada

Mail: ****************@*****.*** LinkedIn: Shivam Ardeshna GitHub: Shivam Ardeshna

EDUCATION

Mila - Quebec AI Institute, Montréal (Affiliated with Université de Montréal) Sep 2025 - Aug 2027

Master’s in Computer Science, Specialization in Machine Learning

Relevant Coursework : Machine Learning, Representation Learning, Data Science

Charotar University of Science and Technology, Gujarat, India Aug 2021 - May 2025

Bachelor’s in Computer Science

Relevant Coursework : NLP, Computer Vision

TECHNICAL SKILLS

Programming Python, C/C++, JavaScript (React, Node.js)

Frameworks PyTorch, TensorFlow, FastAPI, Django, Docker, Kubernetes, OpenCV

Cloud AWS (EC2, SageMaker), GCP, Azure, MLFlow, Lambda Labs

LLM LangChain, LangGraph, vLLM, Ollama, HuggingFace, Weaviate, LanceDB, Pinecone, Qdrant

Tools Git, REST APIs, SQL, Data Visualization, Power BI, Tableau

EXPERIENCE

Machine Learning Engineer May 2024 – May 2025

Digital Hercules Innovations, Noida, India

• Optimized Amazon Ads bidding strategies using ML, increasing ROI by 18% and reducing cost-per-click by

12%.

• Deployed scalable ML pipelines on AWS, reducing model inference time by 35% and boosting data automation

throughput by 40%.

• Led a team of 3 to design AI-powered chat systems, automating 60% of support interactions and improving

response time by 25%.

Machine Learning Engineer Intern Apr 2024 – Jul 2024

Agevole Innovation Pvt. Ltd., Ahmedabad, India

• Devised an NLP-based news recommendation system that improved recommendation accuracy by 25% and

boosted user engagement by 20%.

• Collaborated with cross-functional teams to implement real-time performance optimization and A/B testing

for model deployment.

• Applied data preprocessing and vectorization techniques to reduce latency by 30%.

Software Development Engineer May 2023 – Sep 2023

Heliconia Solutions, Rajkot, India

• Created and executed a Python-Flask based file conversion platform, increasing conversion efficiency by 30%

and reducing processing time by 20%.

• Streamlined backend logic and API integration, reducing error rates by 20% and improving user satisfaction

scores by 15%.

PUBLICATIONS

• Improving Item-Based Collaborative Filtering with Vision Transformers to Address Data Spar-

sity Issues — Published in Springer (ICTCS 2024)

Integrated Vision Transformers with collaborative filtering, improving recommendation accuracy by 17% in

sparse-interaction datasets.

• Evaluating the Efficacy of Diverse Classifiers in Fake News Detection: A Comparative Study

— Published in Springer Nature (ICTCS 2024)

Conducted comparative analysis of ML and DL classifiers for fake news detection, improving accuracy and

robustness through linguistic and ensemble methods.

• Enhancing Deep Fake Image Generation and Detection through Transfer Learning — Accepted

at Springer (AICTA 2024)

Proposed a transfer learning-based deepfake detection pipeline, improving feature extraction accuracy by 15%

and classification precision across benchmark datasets.

PROJECTS

StratifyAI — Django, Financial Data APIs, LLMs

• Engineered a real-time financial analytics platform enabling instant company analysis, earnings tracking, and

sentiment insights using LLMs.

• Integrated live stock APIs and data pipelines supporting 500+ companies, reducing data latency by 40%.

• Enhanced data visualization and KPI dashboards, improving analyst efficiency by 35%.

AI Narrator — LLMs, Vision Models, Multimodal AI

• Built a multimodal AI narration system generating live contextual commentary for video streams, improving

comprehension speed by 50%.

• Designed a low-latency inference pipeline processing 30 FPS video, reducing delay by 25%.

Multimodal VideoRAG — Vision-Language Models, LanceDB, LangChain

• Enhanced a retrieval-augmented video analysis tool enabling natural-language queries on video data, improv-

ing retrieval precision by 35%.

• Combined vision and text embeddings for semantic video understanding and context-aware search.

HONORS & AWARDS

• Kaggle Expert (Notebooks) — Recognized for high-quality machine learning notebooks and community

contributions.Kaggle

• Articles — Seeing Is Believing: How Recommendation Systems Are Learning to See and Vision Transformers:

Demystifying the Latest Breakthrough in Image Recognition were featured in AI Advance and The Modern Scientist

magazines.

Contact this candidate