Job Description -
Role: Technical CPG BA
Work Location: Atlanta, GA (Onsite)
Duration: 3 Months (C2H)
Job Description:
Support the design, development, and validation of an agentic AI solution for R&D teams within a food and beverage organization. This role will operate at the intersection of AI engineering, scientific research, and data modeling, working directly with client scientists and researchers to ensure the solution is grounded in accurate scientific context and delivers reliable, high-quality outputs.
The successful candidate will play a critical role in:
- Structuring domain knowledge into a lightweight ontology to guide agent reasoning
- Defining and curating gold-standard test datasets for evaluating system performance
- Validating and tuning agent outputs to ensure scientific accuracy and usability
This is a hands-on, delivery-oriented role requiring both strong analytical rigor and the ability to engage deeply with scientific stakeholders.
Key Responsibilities:
- Domain Modeling & Ontology Definition in Biochemistry
- Golden Dataset Creation & Evaluation Design
- Agent Validation, QA, and Tuning Support
- Scientific Stakeholder Engagement
- Cross-Source Scientific Data Understanding
Mandatory Skills and Skill Proficiencies Required for This Position:
- Experience working within scientific or R&D environments (food & beverage, biochemistry, or similar preferred) 4
- Experience as a Technical Business Analyst / Product Analyst / Data Analyst in complex systems 3
- Ability to structure ambiguous domains into clear models, schemas, or ontologies 3
- Experience working with GenAI systems (RAG, agents, or similar) 2
- Ability to understand and communicate scientific concepts effectively with SMEs 4
Optional Skills and Skill Proficiencies for This Position:
- Experience with ontologies, knowledge graphs, or semantic layers 3
- Background in chemistry, biochemistry, food science, or related domain 4
- Experience building or evaluating LLM-based applications or agent frameworks 2
- Familiarity with biochemistry and R&D external datasets such as PubMed, Reactome, etc. 3
- Experience defining test strategies for AI systems 2