Data Scientist

Role Summary

AI71 is seeking a Senior Data Scientist to lead the technical execution of high-impact AI initiatives for the EDGE Group. In this role, you will move beyond experimental modeling to build production-grade systems that directly support the design, manufacturing, and procurement of advanced defense capabilities.

You will act as the technical bridge between unstructured data (regulatory text, technical documentation) and structured engineering systems (BIM/IFC models, SAP S/4HANA). Working within a structured "Sprint Zero" to "Stage Gate" delivery model, you will design and deploy the AI engines that power two flagship programs: LeverEDGE (accelerating the systems engineering lifecycle) and Intelligent Supply Chain (predictive spend and risk analytics).

Key Responsibilities

• Generative AI & NLP for Engineering (LeverEDGE)

• Data Exploration and Analysis: Query and analyse large domain- or topic-specific data sets from both structured and unstructured sources, identify patterns and features. Ensure data meets quality standards and requirements before model development.

• Regulation Text Interpretation: Design and fine-tune Large Language Models (LLMs) to parse complex regulatory texts (e.g., building codes, military standards) and extract structured rules for automated compliance checking.

• Rule Formalization: Convert interpreted regulations into computer-processable formats (e.g., object-property-condition-value tuples) that can be executed by downstream compliance engines.

• Querying via NLP: Architect methods for LLMs to map natural language requirements directly to specific metadata entities within various schemas (e.g., mapping "systems design" to specified attributes).

• RAG Architecture: Implement Retrieval-Augmented Generation (RAG) pipelines that allow systems to query vast repositories of technical documentation and historical project data with high accuracy and low hallucination rates.

• Predictive Modeling & Optimization (Supply Chain)

• Forecasting Engines: Develop time-series forecasting models to predict spend categories and material demand by correlating internal ERP data with external macroeconomic signals.

• Classification & Risk Scoring: Build machine learning classifiers to categorize supplier risks and operational anomalies, integrating data from diverse sources to create dynamic risk scores.

• Data Extraction Pipelines: Design robust pipelines to extract and transform raw data (from Data Lakehouse, external web sources, or SAP and other databases) into features required for predictive modeling and automated rule checking. Liaise with Data Engineers on

• System Integration & Performance

• Model Orchestration: Work with Back End Engineers to integrate AI models into a cohesive "compliance engine" or "risk engine" that can be invoked programmatically via robust APIs.

• Optimization: Streamline model performance to ensure complex checks (e.g., analyzing large datasets or processing thousands of supplier records) can be executed within reasonable timeframes, potentially using batching or asynchronous processing.

• Quality Assurance: Validate model outputs against known test cases and historical data, debugging false positives/negatives to refine algorithms and ensure "defense-grade" reliability.

Technical Requirements

• Core AI/ML: Expert proficiency in Python and standard ML libraries (TensorFlow/PyTorch, Scikit-learn, Pandas, NumPy). Strong grasp of both supervised and unsupervised learning techniques.

• NLP & LLMs: Deep experience with transformer-based models (GPT, BERT, Llama) and prompt engineering techniques (few-shot learning, fine-tuning) for domain-specific tasks.

• Data Engineering: Proficiency in handling complex data structures (JSON, XML) and familiarity with database querying (SQL/NoSQL) or graph data structures. Experience with data extraction from specialized formats is a significant plus.

• Backend Awareness: Understanding of how to expose models via RESTful APIs (Flask/FastAPI) and integrate them into larger software architectures.

• Statistics: Solid understanding of statistics, probability distribution, A/B testing. Adept at identifying and mitigating biases in datasets.

Professional Qualifications

• Experience: 5+ years of experience in Data Science or Machine Learning, with a proven track record of deploying models into production environments.

• Domain Adaptability: Ability to quickly grasp complex domain terminology (e.g., construction regulations, defense standards, supply chain taxonomies) and translate them into logic.

• Structured Delivery: Experience working in structured delivery models (Agile/Sprint-based) while adhering to rigorous validation and verification standards.

• Collaboration: Strong communication skills to work effectively with Domain Experts, Backend Engineers, and Product Managers to align model outputs with real-world business logic.

Why This Role?

This is not just about building models; it is about building the intelligence that drives the industrial backbone of the future. You will be solving tangible, high-stakes problems —from ensuring product design and manufacturing being safe and compliant -, to predicting supply chain disruptions before they impact national security capabilities. Join us to turn data into decision advantage.