Dear Hiring Manager, I am interested in the Data and Evidence position. I design and deliver production-grade data science solutions using traditional machine learning alongside modern GenAI and emerging agentic technologies, including using multi-agent systems and MCP to help automating data science tasks: from data cleaning, analysis, modellling and visualization. [3] [4] Beyond standard RAG, I have hands-on experience fine-tuning models like Gemma 3 to meet specific domain requirements. By utilizing Unsloth and LoRA, I’ve streamlined the fine-tuning process to be viable on single-GPU setups without sacrificing model quality. I bridge the gap between training and production by deploying these models via vLLM on AWS, ensuring the final solution is both scalable and cost-effective. My work spans end-to-end data science web apps across time series, geospatial data, images, video, and text, translating complex, real-world data into actionable insights. My online CV with project links is available here: _Website not available. Sign in: https://www.twine.net/signup_ For GenAI agentic applications, I build systems where agents dynamically select prompts, tools, and models based on task context. In multi-agent solutions, each agent has clear role separation (e.g., analyst, coordinator), with robust tool orchestration, memory management, and structured reasoning pipelines such as ReAct-style workflows to ensure reliable and interpretable behaviour as they scale. I am an AWS Certified AI Practitioner [1] as well as Cloud Practitioner [2] with deep understanding of modern neural network architectures, including Transformers, LSTMs, and CNNs, besides classical statistical/ machine learning models. This foundation allows me to select appropriate approaches across predictive modelling, representation learning, and GenAI-driven systems. I follow strong software engineering practices and regularly work with Evidently to monitor data drift, DVC for auto pipelines, Git and GitLab for version control and code review, Docker for reproducible environments, and Nginx + FastAPI for deploying and serving production applications. These practices support secure, maintainable, and scalable AI systems from experimentation through to production. Key Highlights of my work using Python, JavaScript and R: AI Data Scientist, multi-agent system [3] [4] including statistical analysis, data visualization agents, math agent, beside main agent Agentic RAG: with source-validation [5], with tool use and structured output [6]…Dear Hiring Manager, I am interested in the Data and Evidence position. I design and deliver production-grade data science solutions using traditional machine learning alongside modern GenAI and emerging agentic technologies, including using multi-agent systems and MCP to help automating data science tasks: from data cleaning, analysis, modellling and visualization. [3] [4] Beyond standard RAG, I have hands-on experience fine-tuning models like Gemma 3 to meet specific domain requirements. By utilizing Unsloth and LoRA, I’ve streamlined the fine-tuning process to be viable on single-GPU setups without sacrificing model quality. I bridge the gap between training and production by deploying these models via vLLM on AWS, ensuring the final solution is both scalable and cost-effective. My work spans end-to-end data science web apps across time series, geospatial data, images, video, and text, translating complex, real-world data into actionable insights. My online CV with project links is available here: _Website not available. Sign in: https://www.twine.net/signup_ For GenAI agentic applications, I build systems where agents dynamically select prompts, tools, and models based on task context. In multi-agent solutions, each agent has clear role separation (e.g., analyst, coordinator), with robust tool orchestration, memory management, and structured reasoning pipelines such as ReAct-style workflows to ensure reliable and interpretable behaviour as they scale. I am an AWS Certified AI Practitioner [1] as well as Cloud Practitioner [2] with deep understanding of modern neural network architectures, including Transformers, LSTMs, and CNNs, besides classical statistical/ machine learning models. This foundation allows me to select appropriate approaches across predictive modelling, representation learning, and GenAI-driven systems. I follow strong software engineering practices and regularly work with Evidently to monitor data drift, DVC for auto pipelines, Git and GitLab for version control and code review, Docker for reproducible environments, and Nginx + FastAPI for deploying and serving production applications. These practices support secure, maintainable, and scalable AI systems from experimentation through to production. Key Highlights of my work using Python, JavaScript and R: AI Data Scientist, multi-agent system [3] [4] including statistical analysis, data visualization agents, math agent, beside main agent Agentic RAG: with source-validation [5], with tool use and structured output [6]WWWW…