AI Researcher (2 Yr Contract)

Summary This is an AI researcher role. The researcher will design and build non-deterministic analytical tools — grounded in statistical modelling, predictive models, and machine learning — that work across structured and unstructured data. Another part of the role is developing AI agents powered by large language models (LLMs) that can reason over and orchestrate these tools to solve real analytical problems end to end. The researcher is expected to be rigorous with data, curious about intelligent systems, and motivated by seeing their work produce something that actually runs. Responsibilities Research & Analysis • Work with structured datasets to identify patterns, trends, and statistically significant findings that inform research questions and business decisions. • Apply and adapt predictive models (regression, classification, time series forecasting) and mathematical models to address defined research problems. • Conduct exploratory data analysis (EDA) to assess data quality, distributions, and feature relevance prior to modelling. • Design and run statistical experiments, including hypothesis testing, power analysis, and significance testing. • Validate model performance using appropriate evaluation frameworks and communicate limitations honestly. AI development • Study and apply LLM-based agent frameworks to automate and orchestrate research workflows. • Design and build AI agents that use LLMs as a reasoning engine, with data science models and pipelines exposed as callable tools. • Integrate agent outputs with structured data — enabling agents to query databases, invoke predictive models, and interpret results. • Evaluate agent performance: task completion, reasoning coherence, tool-use accuracy, hallucination rate, and failure handling. • Stay current with the evolving agent landscape and share learnings with the team. Technical development • Write clean, well-documented Python code for data processing, feature engineering, model development, and result visualisation. • Query and manipulate structured data from relational databases using SQL. • Use core data science libraries including Pandas, NumPy, Scikit-learn, Matplotlib/Seaborn, etc. • Contribute to shared codebases, following team coding standards and version control practices (KLASS Gitlab). • Assist in building reproducible data pipelines and experiment tracking workflows. Collaboration & communication • Work closely with senior researchers and principal researchers to scope, plan, and deliver project workstreams. • Contribute to the quarterly and annual research reports, translating technical findings into clear, accessible language. • Present findings to internal and client-side collaborators, including non-technical audiences. • Participate in team research reviews, peer code reviews, and knowledge-sharing sessions. Requirements Education • Master’s or PhD – Computer Science, Mathematics, Statistics, Physics, Engineering, Data Science, or a closely related quantitative discipline. (Preferred) Technical skills — essential • Python programming: proficiency with core data science libraries (Pandas, NumPy, Scikit-learn, SciPy, Matplotlib, Seaborn). • Data wrangling: handling missing data, outlier detection, feature engineering, and normalisation on structured datasets. • Mathematical foundations: linear algebra, calculus, and optimisation as applied to machine learning and statistical modelling. • Working knowledge of modelling families: supervised, unsupervised, and generative models Personal Qualities • Research mindsetIntellectually curious across both quantitative and AI domainsRigorous and detail-oriented in analysis and documentationComfortable with ambiguity and iterative problem-solvingHonest about uncertainty — in models and in LLM outputs • Working styleClear communicator, written and verbalCollaborative team player in a research environmentSelf-directed with good time managementEager to learn and receive constructive feedback Experience 0–2 years of professional or research experience We define experience broadly. Relevant experience includes academic research projects, dissertations with a data science component, internships, industry placements, personal or open-source projects, Kaggle competitions, or research assistant roles. We are more interested in the quality of your analytical thinking than the number of years on your CV.