Open to opportunities

PRAVEEN SINGH

@praveensingh1

Message

Experienced Machine Learning Engineer specializing in LLMs and MLOps.

United Arab Emirates

Message

What I'm looking for

I am looking for a role that fosters innovation and collaboration, where I can leverage my skills in machine learning and MLOps to drive impactful projects and contribute to a dynamic team.

I am a Machine Learning Engineer with over 5 years of experience in developing and deploying large language models (LLMs) and multimodal models across various cloud platforms. My expertise lies in fine-tuning large-scale models, building robust ML infrastructure, and optimizing inference performance. I have a strong background in MLOps, cloud deployment, and real-time data processing, which has enabled me to deliver impactful solutions in the field of geotechnical data extraction and analysis.

Throughout my career, I have led the design and development of scalable ML pipelines, achieving significant improvements in accuracy and efficiency. For instance, at Fugro, I fine-tuned multimodal LLMs, increasing accuracy by 30% while reducing compute costs by 28%. I am passionate about mentoring and have successfully guided junior engineers in curating production-ready datasets, ensuring data integrity and usability for model training. My commitment to innovation is reflected in my patent-pending work on an AI-powered Borehole Digitiser, which automates geotechnical data extraction, reducing consultant workload by 90%.

Experience

Work history, roles, and key accomplishments

Current

Machine Learning Engineer

Current

Fugro

Mar 2024 - Present (1 year 4 months)

Fine-tuned multimodal LLMs (Llama3.2-vision, LLaVA2.6, MiniCPM) with Unsloth, DeepSpeed, and LoRA, increasing accuracy by 30% and cutting compute costs by 28% via Amazon SageMaker. Led end-to-end design and development of a scalable ML pipeline leveraging Sonnet3.7 MLLM, achieving 92% accuracy in extracting and structuring borehole log PDFs with graphs using Amazon Bedrock.

LLMs)Amazon SageMaker DeepSpeed MLFlow

Data Scientist

Prospect33.ai

Aug 2023 - Present (1 year 11 months)

Collaborated on building a robust Python function for handling large-scale Financial datasets. Spearheaded feature engineering on extensive merger and acquisition datasets, implementing custom imputation techniques such as DNN and Midaspy.

Python FairCutForest Rough K Means Sagemaker

Machine Learning Engineer

Sponsorlytix Ltd.

May 2022 - Present (3 years 2 months)

Engineered custom yolov5 models in Python (PyTorch) as a computer vision pipeline, transforming vast amounts of data for different sports Ads and sponsor logos with Roboflow. Accelerated hyperparameter tuning, experiment tracking and model registry using MLflow, ensuring precise monitoring of production model drifts and wrapped the prediction service in Flask API, reducing latency by an impressive

YOLOv5 PyTorch Computer Vision Roboflow MLFlow Flask API AWS Docker YOLO V5

Senior Software Engineer

MakeMyTrip Ltd.

Jun 2019 - Present (6 years 1 month)

Implemented a MongoDB-backed GraphQL content delivery system, integrating real-time COVID-19 regulation updates with 120ms response time, serving 2000 million requests/month. Built a Python-Django-based flight data API, cutting manual configuration errors by 75%, enhancing decision-making for 11 airline partners.

MongoDB GraphQL Python Django Flask REST APIs Automation API Development

Intern

Defence Research and Development Organisation

Jun 2018 - Present (7 years 1 month)

Built a Ship Recognition System using TensorFlow Object Detection API. Trained Faster-RCNN model with up to 99.5% accuracy in image classification and localization of different ships in a video flink.

TensorFlow Object Detection Faster RCNN

Intern

Risk Latte Americas Inc.

Dec 2017 - Present (7 years 7 months)

Researched and analyzed various machine learning algorithms in computational mathematics. Coded Genetic Algorithm in Excel VBA for solving Ordinary differential equations with an accuracy of 98%.

Machine Learning Algorithms Genetic Algorithms VBA Ordinary Differential Equations Ordinary Differential Equation

Education

Degrees, certifications, and relevant coursework

Concordia University

Master of Applied Computer Science, Applied Computer Science

Completed a Master of Applied Computer Science, gaining advanced knowledge in various computing disciplines. Focused on practical applications and cutting-edge technologies in the field.

Delhi Technological University

Bachelor of Technology, Mathematics Computing

Activities and societies: Organized the MOCK STOCK event under SIAM – Society for Industrial and Applied Mathematics; Served as a volunteer in the Hospitality Team for ENGIFEST 2016 - DTU University cultural festival.

Earned a Bachelor of Technology in Mathematics Computing, building a strong foundation in computational mathematics and technology. Engaged in various academic and extracurricular activities, including organizing events and volunteering.