Principal AI Solution Engineer / AI Architect
Principal AI Solution Engineer / AI Architect
Syneos Health® is a leading fully-integrated life sciences services organization built to accelerate customer success. We partner with innovators at every point across the drug development and commercialization continuum, helping them navigate complexity, anticipate change and accelerate progress.
Every day we perform better because of how we work together, as one team, each the best at what we do. We bring together talented experts across a broad spectrum of business critical corporate functions. Every role plays an essential part in enabling our customers to achieve their goals. Our teams are agile, collaborative, and committed to delivering—for each other, for our customers, and ultimately for the people who rely on the services we support.
Discover what your 25,000 future colleagues already know
Why Syneos Health
- We are passionate about developing our people, through career development and progression; supportive and engaged line management; technical and therapeutic area training; peer recognition and total rewards program.
- We are committed to building an inclusive culture – where you can authentically be yourself. Central to this is our purpose – Driven to Deliver – which captures the passion of our colleagues to show up each day and shape solutions that have the ability to dramatically impact someone’s life.
- We are continuously building the company we all want to work for and our customers want to work with. Why? Because we know that when we bring together smart colleagues from across the world, we can shape the future of healthcare, driving impact for customers and defining the pace of patient progress.
Job Responsibilities
Role Summary
We are seeking a Principal AI Solution Engineer / AI Architect with a minimum of 8 years of overall technology experience, a strong foundation in data science and machine learning, and a record of designing and delivering multiple production-grade AI solutions. This is a hands-on technical leadership role responsible for understanding business requirements, determining whether each problem is best solved with machine learning, generative AI, agentic AI, deterministic logic, or a combination of these, architecting the solution accordingly, and verifying and validating that the implemented solution meets its requirements before and after production deployment.
The ideal candidate combines solid grounding in statistics, machine learning, and model evaluation with deep expertise in Azure AI Foundry, Azure OpenAI, Azure AI Search, retrieval-augmented generation (RAG), and agent-based systems, along with strong enterprise integration skills across REST APIs, Oracle Database, Oracle APEX, and the Microsoft ecosystem. The role requires sound architectural judgement, the discipline to choose the simplest effective approach, practical engineering depth, stakeholder influence, and the ability to establish reusable patterns for responsible AI adoption.
Key Responsibilities
- Own the end-to-end architecture and technical delivery of enterprise AI solutions, from requirement discovery and proof of concept through verification, validation, production rollout, and continuous improvement.
- Lead requirement analysis with business and domain stakeholders to understand the problem, users, data, constraints, risks, and measurable outcomes, and translate them into clear functional, non-functional, and acceptance criteria.
- Assess each use case and determine whether it is best served by traditional machine learning, generative AI, agentic AI, deterministic rules, or a hybrid approach, documenting the rationale, trade-offs, and risks behind each architecture decision.
- Partner with business, product, data, security, infrastructure, and engineering stakeholders to identify and prioritize high-value use cases and align solutions with enterprise architecture, data, and security standards.
- Architect machine learning solutions with data science teams, covering problem formulation, data and feature pipeline design, model selection, training and validation strategy, deployment, monitoring, and retraining.
- Design secure, scalable, reliable, and cost-conscious solutions using Azure AI Foundry, Azure OpenAI, Azure AI Search, and appropriate agent frameworks.
- Architect RAG solutions, including document ingestion, chunking, embeddings, vector indexing, hybrid retrieval, reranking, grounding, citation, and response-generation strategies.
- Design AI agents and agentic workflows using tool calling, orchestration, state and memory patterns, guardrails, exception handling, and human-in-the-loop approvals.
- Integrate AI capabilities with enterprise applications and data sources through REST APIs, Oracle Database, Oracle APEX, and Microsoft ecosystem services.
- Establish reference architectures, reusable components, engineering standards, prompt patterns, evaluation methods, and deployment practices for AI solutions.
- Define and implement controls for responsible AI, data privacy, security, access management, content safety, auditability, and regulatory or organizational compliance.
- Own verification and validation of implemented solutions, confirming through test plans, requirement traceability, benchmark datasets, statistical evaluation, and user acceptance testing that each solution is built correctly and meets the business need before release.
- Create evaluation frameworks covering model performance, answer quality, retrieval relevance, groundedness, safety, latency, reliability, user adoption, and cost, and use post-deployment monitoring to revalidate and improve solutions iteratively.
- Provide hands-on technical guidance, conduct architecture and code reviews, troubleshoot complex issues, and help teams make pragmatic build-versus-buy decisions.
- Mentor senior engineers, data scientists, and AI champions, accelerate capability development across application teams, and promote disciplined experimentation and knowledge sharing.
- Communicate architecture decisions, trade-offs, delivery risks, and business value clearly to both technical teams and senior leadership.
- Contribute to Agile planning, estimation, dependency management, release readiness, production support, and post-implementation reviews.
Required Skills
AI Platforms and Solution Architecture
- Proven ability to translate business requirements into solution architectures and to select between machine learning, generative AI, agentic, rule-based, and hybrid patterns based on the problem, available data, risk, explainability, and cost.
- Strong hands-on experience with Azure AI Foundry and Azure OpenAI for building, evaluating, deploying, and operating generative AI solutions.
- Experience with Azure AI Search, including vector search, hybrid search, semantic ranking, indexing strategies, filters, and security-aware retrieval.
- Practical experience with agent frameworks and orchestration patterns for enterprise use cases.
- Ability to design cloud-based AI architectures that balance scalability, availability, security, performance, maintainability, and cost.
- Understanding of model selection, model limitations, context management, token usage, latency, throughput, rate limits, and fallback strategies.
Data Science and Machine Learning
- Strong foundation in statistics, probability, experimental design, and hypothesis testing, applied to model evaluation and decision-making.
- Hands-on experience with supervised and unsupervised learning, including classification, regression, clustering, time-series forecasting, and anomaly detection, using libraries such as scikit-learn, XGBoost, PyTorch, or TensorFlow.
- Solid understanding of data quality, feature engineering, training and validation strategies, overfitting, data leakage, model calibration, and performance measures suited to the business problem.
- Ability to interpret model outputs, explain model behavior using explainability techniques, and assess bias and fairness.
- Experience with MLOps practices, including experiment tracking, model registries, model monitoring, drift detection, and retraining pipelines.
- Clear judgement on when classical machine learning or deterministic approaches are more appropriate, reliable, or cost-effective than generative AI or agentic solutions.
Retrieval-Augmented Generation (RAG)
- Deep understanding of embeddings, vector databases, retrieval architectures, and grounding strategies.
- Experience designing ingestion and retrieval pipelines for structured and unstructured enterprise content.
- Knowledge of chunking, metadata design, query transformation, hybrid retrieval, reranking, citation, access controls, and freshness strategies.
- Ability to evaluate and improve retrieval relevance, answer groundedness, completeness, and hallucination risk.
Agent Development
- Understanding of single-agent and multi-agent workflows, agent orchestration, planning, routing, state, memory, and task decomposition.
- Experience with function or tool calling, structured outputs, API-based actions, error recovery, and observability.
- Ability to design human-in-the-loop approval and escalation controls for sensitive or high-impact actions.
- Knowledge of guardrails that prevent unauthorized, unsafe, or unintended agent behavior.
Prompt Engineering, Evaluation, and Validation
- Advanced prompt engineering skills, including system instructions, few-shot prompting, decomposition, structured output, grounding, and reusable prompt templates.
- Experience designing automated and human evaluation approaches, test datasets, quality rubrics, regression tests, and production feedback loops.
- Experience defining verification and validation plans for machine learning, generative AI, and agentic solutions, including requirement traceability, acceptance criteria, test coverage, benchmark datasets, and sign-off evidence.
- Ability to apply statistical rigor to...