Data Scientist

About the Role

We are seeking a Data Scientist to help analyze, improve, and evolve existing codebases and data-processing pipelines into robust, scalable, and maintainable solutions.

In this role, you will work closely with our data and application development teams to develop practical AI and machine learning solutions that support Irth’s core business areas, including Damage Prevention, Asset Integrity, Land Management, and Stakeholder Engagement.

The ideal candidate is comfortable working across data exploration, machine learning, NLP/LLM experimentation, software development, and productionization.

Key Responsibilities

• Design, prototype, and evaluate LLM/NLP solutions to test hypotheses and address business and technical needs, including:
• Text classification
• Entity extraction
• Semantic search
• Summarization
• Other natural language processing and AI applications
• Translate successful prototypes into production-ready solutions using clean, modular, maintainable, and well-tested code.
• Apply sound object-oriented design and software engineering practices when developing and restructuring solutions.
• Analyze and understand existing codebases, data pipelines, and applications to determine what should be preserved, improved, consolidated, or redesigned.
• Explore, prepare, transform, and analyze data independently using Databricks to support machine learning models and data-processing pipelines.
• Collaborate closely with the data engineering and application development teams to ensure effective integration between Databricks, data pipelines, models, and application services.
• Participate in technical discussions and architectural decisions, providing recommendations for effective and maintainable solutions.
• Help establish development practices that improve the reliability, scalability, testing, and maintainability of data science solutions.
• Work collaboratively with engineering, data, and leadership teams to translate business requirements into practical technical solutions.

What Success Looks Like

In this role, you will help Irth move AI and data science initiatives from early-stage experimentation to reliable production solutions. You will contribute to building scalable data pipelines, reusable machine learning capabilities, and maintainable software that delivers measurable value across Irth’s products and industries.

Requirements

Required Qualifications

• 3–5 years of experience in data science, machine learning, or ML engineering, with demonstrated experience developing production-quality code beyond notebooks.
• Master’s degree or equivalent professional experience in data science, machine learning, computer science, statistics, or a related field, with a strong foundation in statistics and modeling methodologies.
• Hands-on experience developing NLP and/or LLM-based solutions, including:
• Retrieval-Augmented Generation (RAG)
• Embeddings and similarity search, including cosine similarity
• Hybrid search combining BM25 and semantic search
• Text deduplication techniques such as fuzzy matching and MinHash
• Text classification and Named Entity Recognition (NER)
• Search re-ranking
• Prompt engineering
• Strong proficiency in Python, including experience with pandas, scikit-learn, PyTorch, and other relevant machine learning libraries.
• Strong understanding of object-oriented programming, clean code, and software design principles.
• Strong data analysis skills with experience applying statistical methods such as exploratory data analysis, hypothesis testing, and statistical modeling.
• Hands-on experience with Databricks or a comparable data/analytics platform, including the ability to independently explore, transform, and prepare data.
• Solid understanding of the machine learning lifecycle, including model training, evaluation, deployment, and production monitoring for issues such as model drift and data quality.
• Experience designing and developing REST APIs to expose data, models, or analytical results to downstream applications.
• Strong written and verbal communication skills, with the ability to collaborate effectively across technical teams with different responsibilities and priorities.

Nice-to-Have Qualifications

• Experience working with news, social media, or other large-scale unstructured text data, including media monitoring, web scraping, or APIs.
• Domain knowledge or professional experience in communications, public relations, media, or related industries.
• Familiarity with the broader LLM ecosystem, including vector databases/stores, RAG frameworks, and fine-tuning NLP models.
• Experience with MLflow or similar experiment-tracking and model-management tools.
• Basic knowledge of TypeScript and Node.js.
• Experience with CI/CD practices and version control, including Git, GitHub Actions, or equivalent tools.