Artificial Intelligence Engineer

About the Company

Trust3 AI is an AI platform offered by Privacera. The flagship product Trust3 IQ is the Universal Context Engine that makes AI agents enterprise-grade by creating, serving, and continuously improving canonical business context and governance for downstream agents. Trust3 IQ supplies semantically rich, audited context to agents via the Model Context Protocol (MCP), enforces governance at runtime (with integration with Privacera’s access governance), and measures agent trust and visibility. We’re hiring ML/AI engineers to turn that context into reliable, governed agent behavior used by enterprise customers.

About the Role

We’re seeking a hands-on AI Engineer (Mid → Senior) with strong ML/AI experience to design, build, and productionize LLMs, RAG systems, Data/SQL agents and evaluation pipelines that consume Trust3 IQ context (via MCP). This role focuses on model engineering, evaluation, MLOps and trust/visibility metrics as part of our core engine work.

Responsibilities

• Own design, training and fine-tuning of LLM/NLP models and components used by Trust3 IQ agents (instruction tuning, LoRA/adapters, embeddings).
• Design and implement ML/agent logic for Data/SQL agents (text-to-SQL, plan generation, multi-step agent orchestration) that leverage MCP context and meet enterprise accuracy requirements.
• Build robust retrieval / RAG architectures: embedding pipelines, vector DB strategies, hybrid retrieval and prompt orchestration tuned to IQ context.
• Integrate with MCP as a context consumer — design context-aware prompts and caching strategies to consume entity definitions, metric formulas and policy flags.
• Create and own LLM/agent evaluation suites (automated + human-in-the-loop), define acceptance thresholds and push evaluation into CI.
• Implement governance-aware modeling: ensure models and retrieval honor runtime policies (access/masking/redaction/deny) and design safe fallbacks or explanations.
• Define and implement Trust scores and telemetry for agent interactions; instrument pipelines for drift detection, audit trails and continuous learning from SME feedback.
• Productionize model infra: training pipelines, versioning, model serving, monitoring (quality, latency, cost), and SLOs.
• Use modern MLOps tooling (MLflow, W&B, Arize).

Qualifications

• 3+ years of professional experience building production ML/AI systems with hands-on work in LLM/NLP (fine-tuning, embeddings, RAG).
• Strong PyTorch experience (or equivalent) and solid ML engineering skills (data pipelines, experiments, reproducibility).
• Practical experience building retrieval/RAG pipelines and working with vector DBs.
• Experience designing LLM/agent evaluation suites and interpreting metrics.
• Strong software engineering fundamentals: testing, CI/CD, version control, containerization and cloud deployments.
• Excellent written and verbal communication — able to explain model behavior and tradeoffs to product and customer stakeholders.

Preferred Skills

• Experience architecting agentic systems, text-to-SQL or SQL/BI agent design and benchmarking (bird-bench or similar).
• Knowledge or hands-on experience designing eval metrics and human-in-the-loop learning systems.
• Familiarity with knowledge graphs / semantic inventories and how to use them as prompt/context sources.
• MS/PhD in Machine Learning, NLP, Computer Science, or a related field (preferred for senior levels).

Equal Opportunity Statement

Privacera / Trust3 IQ is an equal opportunity employer. We celebrate diversity and are committed to creating an inclusive environment for all employees.