Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi, I’m Harsh Tyagi, an AI/ML Engineer based in Dublin, Ireland, with 2+ years of experience building and deploying deep learning models for anomaly detection, fraud risk scoring, and predictive analytics. I’m adept at turning complex data into scalable ML solutions that drive measurable security and business impact.\n\nI specialize in PyTorch, ONNX, MLflow, SHAP, and Evidently, with hands-on experience in LSTM architectures, real-time inference, and model governance. I’m passionate about building explainable, high-accuracy AI systems that integrate seamlessly into production pipelines and foster collaboration with cross-functional teams.…Hi, I’m Harsh Tyagi, an AI/ML Engineer based in Dublin, Ireland, with 2+ years of experience building and deploying deep learning models for anomaly detection, fraud risk scoring, and predictive analytics. I’m adept at turning complex data into scalable ML solutions that drive measurable security and business impact.\n\nI specialize in PyTorch, ONNX, MLflow, SHAP, and Evidently, with hands-on experience in LSTM architectures, real-time inference, and model governance. I’m passionate about building explainable, high-accuracy AI systems that integrate seamlessly into production pipelines and foster collaboration with cross-functional teams.

Harsh Tyagi

Data Analyst, Data Scientist, Web Developer, +4





Hi, I’m Harsh Tyagi, an AI/ML Engineer based in Dublin, Ireland, with 2+ years of experience building and deploying deep learning models for anomaly detection, fraud risk scoring, and predictive analytics. I’m adept at turning complex data into scalable ML solutions that drive measurable security and business impact.\n\nI specialize in PyTorch, ONNX, MLflow, SHAP, and Evidently, with hands-on experience in LSTM architectures, real-time inference, and model governance. I’m passionate about building explainable, high-accuracy AI systems that integrate seamlessly into production pipelines and foster collaboration with cross-functional teams.…Hi, I’m Harsh Tyagi, an AI/ML Engineer based in Dublin, Ireland, with 2+ years of experience building and deploying deep learning models for anomaly detection, fraud risk scoring, and predictive analytics. I’m adept at turning complex data into scalable ML solutions that drive measurable security and business impact.\n\nI specialize in PyTorch, ONNX, MLflow, SHAP, and Evidently, with hands-on experience in LSTM architectures, real-time inference, and model governance. I’m passionate about building explainable, high-accuracy AI systems that integrate seamlessly into production pipelines and foster collaboration with cross-functional teams.

Available to hire

Hi, I’m Harsh Tyagi, an AI/ML Engineer based in Dublin, Ireland, with 2+ years of experience building and deploying deep learning models for anomaly detection, fraud risk scoring, and predictive analytics. I’m adept at turning complex data into scalable ML solutions that drive measurable security and business impact.\n\nI specialize in PyTorch, ONNX, MLflow, SHAP, and Evidently, with hands-on experience in LSTM architectures, real-time inference, and model governance. I’m passionate about building explainable, high-accuracy AI systems that integrate seamlessly into production pipelines and foster collaboration with cross-functional teams.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

English

Fluent

Work Experience

Data Scientist at K2 Analytics

June 1, 2024 - June 1, 2024

Designed and deployed financial and pricing models for credit scoring, churn prediction, and fraud detection using Python and Scikit-learn — improving prediction accuracy by 85% and reducing false positives by 30%. Designed and automated ETL/ELT and feature-engineering pipelines in Python, PySpark and SQL, reducing data preparation time by 40% and ensuring clean, reliable datasets for modeling. Delivered 100+ interactive dashboards in Power BI and Tableau, applying DAX, modeling, and performance tuning to streamline reporting for compliance, pricing, and operational teams. Partnered with stakeholders to identify key metrics, visualize trends, and present actionable insights, driving measurable improvements in operations and reporting efficiency. Conducted sensitivity and scenario analyses on revenue and loss models using Excel and Python to optimize pricing strategies and support data-driven underwriting decisions.

Software Engineer at AdGlobal360

April 1, 2022 - April 1, 2022

Boosted website performance by 15% through enhanced component development and content optimization using Sitecore, .NET, and object-oriented programming principles. Accelerated delivery cycles by 20% by developing modular code and aligning with Agile SDLC workflows.

Data Intern at Grass Solution

June 1, 2019 - June 1, 2019

Built image classification and spam detection models with 90%+ accuracy using Python and Scikit-learn, automating manual review processes. Conducted data preprocessing, validation, and statistical analysis on unstructured datasets, improving scalability by 20%.

Data Scientist at K2 Analytics

April 1, 2022 - June 1, 2024

Designed and deployed predictive ML models for credit scoring, churn prediction, and fraud detection using Python and scikit-learn, improving prediction accuracy and reducing false positives. Led A/B testing and experiment design, driving ROI uplift. Engineered ETL/ELT pipelines in Python, SQL, and PySpark to streamline data prep and improve model stability. Delivered 100+ dashboards (Power BI, Tableau) for company-wide insights and translated predictive insights into actionable strategies. Implemented data quality checks and governance to improve data lineage and audit readiness across modelling pipelines.

Software Engineer at AdGlobal360

May 1, 2021 - April 1, 2022

Boosted website performance by 15% through modular component development and content optimization; accelerated delivery cycles by improving code practices and aligning with Agile SDLC.

Sales Associate at TKMaxx

November 1, 2024 - Present

Demonstrated reliability and customer service skills while managing inventory and assisting shoppers in a fast-paced retail environment during postgraduate studies.