Pattern Agent Developer Job Phoenix area,Arizona USA,Software Development

Job Summary

We are looking for a highly skilled ML / AI Engineer to design and implement the Pattern Agent algorithm and the Zone 3 write-path architecture. The ideal candidate will have strong expertise in machine learning pipelines, probabilistic models, knowledge graph systems, and retrieval architectures. This role focuses on building intelligent pattern validation, correction capture, and graduation systems for a shared knowledge layer.

Key Responsibilities

Design and implement the Pattern Agent algorithm, including:

Probabilistic promotion logic
Anonymization rules
Confidence scoring
Graduation criteria for the Zone 3 write-path (D9)

Build the three-layer matching architecture:

Direct lookup (zero token cost, 80%+ classification coverage)

Constrained similarity search against the Zone 3 knowledge graph

Full LLM inference as the final fallback

Develop the Zone 3 write-path, including:

Approval routing for new pattern submissions
Anonymization pipeline for PII and client‑identifying information removal
Write‑gate validation to prevent unvalidated patterns entering the shared layer

Implement correction capture mechanisms to record practitioner approve / modify / reject decisions and feed them into the graduation pipeline.

Design and document the graduation criteria and probabilistic model for promoting client‑specific patterns into shared intelligence.

Collaborate with frontend developers, GraphRAG/Data Engineers, and AI platform teams to integrate pattern intelligence workflows.

Ensure scalability, reliability, and maintainability of ML inference and orchestration pipelines.

Required Skills

Strong expertise in ML pipeline design, including training loops, inference pipelines, and confidence scoring.
Experience with probabilistic/statistical models, such as Bayesian updating or equivalent graduation algorithms.
Knowledge of graph data structures, ontology design, graph traversal, and knowledge graph modeling.
Experience designing RAG retrieval architectures, including chunking strategies and context window management.
Strong programming skills in Python and experience with ML/AI frameworks.
Ability to document technical designs, probabilistic models, and graduation criteria clearly.

Preferred Skills

Experience with knowledge graph platforms such as Neo4j or Amazon Neptune.
Familiarity with Amazon Titan v2 embeddings or equivalent embedding model experience.
Experience with LLM inference and orchestration workflows.
Exposure to enterprise AI systems involving pattern intelligence and retrieval pipelines.

Explicitly Not Required

LLM pre‑training or fine‑tuning
Frontend development
AWS infrastructure provisioning (owned by IaC Architect)
Vector database administration (owned by GraphRAG / Data Engineer)

Preferred Candidate Profile

Strong background in Machine Learning, AI Engineering, Data Science, or a related field.
Experience building production‑grade ML systems and intelligent retrieval architectures.
Ability to work independently and collaborate across engineering, product, and AI teams.
Strong analytical, problem‑solving, and documentation skills.

#J-18808-Ljbffr