AI/ML Engineer, Algorithm

Role Overview

• Design and implement production multi-agent AI systems: agent coordination, tool use, context and memory management, structured output, and human-in-the-loop ratification loops.
• Develop novel algorithms for scoring, ranking, prioritization, and optimization
• including multi-dimensional impact scoring systems that rank ideas or proposals by value, feasibility, revenue potential, and strategic alignment.
• Build LLM-powered workflows: retrieval-augmented generation (RAG), prompt pipelines, fine-tuning, structured extraction, and domain-adapted language models for enterprise contexts.
• Architect closed-loop AI systems where agent observations feed structured proposals for human review and ratification, then feed back into continuous improvement cycles.
• Identify novel algorithmic and AI system designs that represent patentable inventions; author invention disclosures and collaborate with patent counsel through prosecution.
• Lead evaluation and benchmarking of AI model performance in production: accuracy, latency, cost, reliability, and domain-specific accuracy metrics.
• Drive build-vs-buy-vs-partner decisions for AI components; design integration patterns for third-party foundation models, AI APIs, and specialized ML services.
• Establish AI governance practices: model cards, explainability tooling, bias evaluation, auditability, and responsible deployment standards.
• Design AI and agent systems that operate across decentralized, federated data domains
• enabling intelligent query routing, context assembly, and inference across disparate operational data sources without requiring centralized data pipelines; architect retrieval and grounding strategies that respect data sovereignty and domain ownership as the organization transitions from siloed data stores to a federated intelligence layer.

Requirements

• 8+ years in AI/ML engineering or research, with at least 3 years building and operating production AI systems at a senior IC level.
• Deep expertise in large language models and agentic frameworks (LangChain, LangGraph, AutoGen, CrewAI, or equivalent); strong understanding of prompt engineering and model behavior.
• Proven ability to develop novel algorithmic approaches
• not just apply existing frameworks
• and document them rigorously for IP purposes.
• Strong ML fundamentals: optimization theory, probabilistic modeling, statistical learning, reinforcement learning, and deep learning architectures.
• Proficiency in Python and at least one deep learning framework (PyTorch or JAX); familiarity with ML infrastructure (MLflow, Weights & Biases, Ray, or equivalent).
• Experience with vector databases, embeddings, semantic search, and knowledge graph architectures as components of production AI systems.
• Demonstrated ability to deploy and maintain AI systems in production: monitoring, versioning, rollback strategies, and performance degradation detection.
• Named inventor on AI, ML, or algorithmic patents (preferred).
• Experience with autonomous AI systems, robotic process automation, or AI applied to physical infrastructure, industrial operations, or IoT environments (preferred).
• Familiarity with RLHF, Constitutional AI, or other alignment and fine-tuning methodologies (preferred).
• Experience with digital twin architectures, simulation environments, or AI-driven 3D modeling tools in operational contexts (preferred).
• Background in algorithm design for scheduling, resource allocation, combinatorial optimization, or operational research (preferred).
• Experience building AI systems that process structured operational data: time-series sensor readings, maintenance logs, energy consumption data, or workforce telemetry (preferred).
• Familiarity with federated learning frameworks or privacy-preserving ML techniques that enable model training across distributed, domain-owned data sources without centralizing sensitive operational data (preferred).
• Experience integrating AI agents with federated data access layers
• query routing across multiple source systems, semantic data discovery, and context-aware retrieval from heterogeneous operational datastores (preferred).
• Graduate degree (MS or PhD) in Computer Science, Machine Learning, Mathematics, or a related quantitative field.

Tech Stack

• Assembly
• IoT
• Python
• PyTorch
• Ray
• RPA

Benefits

• health, dental, and vision insurance
• Health Savings Account (HSA)
• gym discount
• mental health
• Discounted Group Life & AD&D
• Discounted Group Short & Long-term Disability
• 401(k) retirement matching
• PTO/paid holidays
• parental leave