AI/Machine Learning Engineer [Contractor]
WANTED: Thinkers, Builders & Dreamers
For over 15 years, ArcTouch has created lovable apps, websites, and connected experiences for world-class companies, leading brands, and innovative startups. See some of our work.
Since launching our first app in 2009, we’ve followed these core beliefs:
Create lovable products, the perfect blend of something useful and delightful that forges meaningful connections with users.
Be a lovable company where amazing teams of diverse individuals collaborate every day to create something we can be proud of.
ArcTouch is a specialty studio that is part of the award-winning AKQA Group.
If you're passionate about building lovable products, we want you to join us!
Role Summary
We are looking for a Senior AI/ML Engineer to lead the architecture of a next-generation LLM enrichment pipeline within an enterprise Knowledge Graph for a global CPG/Food leader. In this role, you will leverage Microsoft Fabric IQ and Azure OpenAI to transform unstructured product data into high-value, structured metadata at scale (dietary classifications, flavor profiles, and sourcing attributes). You will have end-to-end ownership of taxonomy design, prompt engineering strategies, and evaluation frameworks, ensuring high precision before models are handed off for human-in-the-loop review.
Key Responsibilities:
Design taxonomy for LLM classification (dietary, flavor, use case, sourcing) with the Data Architect and Director of AI
Design and iterate prompts for entity classification and metadata generation across products, recipes, and ingredients
Build an evaluation harness to measure enrichment accuracy, coverage, and drift
Implement the enrichment pipeline in Azure OpenAI with a human review step before writing to the graph
Support entity resolution (matching recipe ingredients to Salsify products, disambiguating variants like ground vs. whole peppercorn) with backend engineering
Define escalation paths for dietary/health claim outputs needing regulatory or Director of AI sign off
Tune model selection and quality thresholds against the review policy
Document taxonomy decisions, prompt versions, and evaluation results
Required Qualifications:
Solid experience in ML/AI engineering with hands-on LLM classification or enrichment pipeline experience
Strong Python; prompt engineering and structured output generation (JSON schemas, function calling)
Hands-on experience with Azure OpenAI or an equivalent LLM API
Experience building evaluation/benchmarking frameworks for model outputs
Experience building human-in-the-loop review workflows
Familiarity with taxonomy or ontology-driven data models
Working knowledge of vector search/embeddings for semantic matching
Nice to have:
CPG, retail, or food industry experience, especially dietary/health claim content
Familiarity with schema.org, JSON-LD, and GEO
Experience with Microsoft Fabric/Fabric IQ
Microsoft certifications: Azure AI Engineer Associate or equivalent
REMOTE WORK: This position is remote-friendly. If you miss being in the office, we have a great one in Florianópolis, and you are welcome to work from there whenever you wish. Just a reminder: remote employees must live in Brazil (we hire international professionals only for specific openings).
Why work for us? Simple: We love what we do.
Make sure to check out our employee reviews on Glassdoor to find out why they love working for ArcTouch.
ArcTouch is an Equal Opportunity Employer (EOE). Applicants are considered for employment without regard to age, race, color, religion, sex, national origin, sexual orientation, gender identity and gender expression, disability, or veteran status.
Come join us!