Principal Data Scientist, SIP

The Advanced Engineering team is Invesco's AI center of excellence, driving innovation through IVYGPT, agentic AI frameworks, and business-sponsored GenAI solutions that improve efficiency across the firm. The team also promotes AI adoption through meetups, hackathons, and cross-functional collaboration.

We are seeking a Principal Data Scientist in Atlanta to design, build, and validate agentic AI systems that solve complex financial challenges. Using tools such as Copilot Studio, LangGraph, Cortex, and AgentCore, you will create multi-agent workflows, develop evaluation frameworks, and lead AI initiatives from concept to production.

This role is ideal for a hands-on leader who thrives on building, testing, and scaling AI solutions.

- Leading and collaborating with software engineers, data engineers, and investment professionals to build innovative AI-based tools using structured and alternate data sources.

- Designing and implementing complex reasoning loops using frameworks like LangGraph, LangChain, or AgentCore, transforming static LLMs into dynamic, goal-oriented agents.

- Developing robust evaluation frameworks and "gold-standard" datasets to measure the performance, faithfulness, and accuracy of generative outputs.

- Scaling AI initiatives and agentic development across Invesco Technology teams by creating reusable patterns, prompt templates, agentic harnesses, skills, and subagents with GitHub Copilot and Claude Code.

- Collaborating with full stack engineers to integrate AI models into business workflows, using AI-assisted coding to move from a notebook experiment to a functional microservice in record time.

- Implementing systematic testing (e.g., RAG evaluation, stress testing) to ensure AI solutions meet the high compliance and reliability standards of the financial services industry.

- Applying emerging techniques in prompt engineering, fine-tuning, and multi-agent systems to solve unique business challenges.

- 7-10 years of demonstrated experience in data science or machine learning, with a strong emphasis on Natural Language Processing (NLP) and Generative AI.

- Strong foundation in Python and cloud-native development; you write clean, modular code and are comfortable with Docker, CI/CD, and DevSecOps.

- Exposure to or high enthusiasm for learning LangGraph, LangChain, Copilot Studio, or AgentCore with Strands.

- Familiarity with creating evaluation metrics and building datasets for model validation and testing.

- Proactive use of AI tools (GitHub Copilot, Cursor, etc.) to accelerate technical tasks and explore new libraries.

- A desire to work at the intersection of data engineering and full-stack development, with the agility to pivot based on rapid sprint feedback.