Executive Director, Applied AI Solution

Job Description:

Position Summary

The Executive Director, Applied AI Solutions, is the senior technical leader responsible for translating ambiguous business problems at New Amsterdam Pharma into working software solutions. The role leads a team of AI solution engineers (whether internal staff or external contributors) and is accountable for the quality, delivery, and evolution of the work. The current technical emphasis is on AI agents, the semantic and metadata layers that make enterprise data legible to those agents, and the workflow automation that connects both to the broader business. The scope is defined by a way of working rather than by ownership of any single system, and the technologies in use are expected to evolve over time.

The role spans the full lifecycle of a solution: identifying where friction exists in a process, designing a tractable approach, building it, deploying it, and iterating on it in production. It draws on a broad technical toolkit and applies it where it has the most leverage.

The Applied Solutions Design Framework

This role is central to NAP's ambition to become an AI-native organization where AI meaningfully accelerates clinical and operational decision-making. The work done here, connecting enterprise data to intelligent systems, reducing friction in high-stakes processes, and building the infrastructure that makes AI agents trustworthy in a regulated environment, directly shapes how NAP operates as it advances through clinical development and toward commercialization.

The role is organized around a recurring five-step loop. Each step represents a distinct discipline.

1. Problem Discovery. Invest the time required to fully understand a problem before committing to a solution. Identify who is affected, what the failure mode looks like in practice, and what a successful outcome would mean for stakeholders. Validate assumptions and challenge pre-supplied answers. Most of a solution’s eventual value originates in the rigor of this step.

2. Solution Design. Translate the problem into a workable architecture: the systems involved, the relevant trust boundaries, the consumer-facing surface, and the elements deliberately left out of scope.

3. Innovation. Apply the most appropriate technology for the problem, drawing from a wide toolkit that may include AI agents and agent infrastructure, semantic data layers, custom applications, media processing, BI tools, or workflow automation. Avoid forcing a problem to fit a familiar tool.

4. Delivery. Build, harden, integrate, deploy, and operate the solution. Carry the work through the full path to production and ongoing use.

5. Iteration. Observe how users actually engage with the solution, identify drift between intent and behavior, and revisit earlier steps as evidence warrants. Iteration is treated as core to the role rather than a follow-on activity.

Core Responsibilities

AI agents and agent enablement

Lead the design and development of AI agents serving internal users at NAP, along with the supporting infrastructure that makes NAP's data and systems accessible to those agents. Current scope includes:

• The unified metadata API and semantic layer connecting clinical-trial, commercial, and enterprise data to agent consumers via MCP (Model Context Protocol, a standard for connecting AI agents to enterprise data services).
• The metadata-generation and enrichment pipeline that powers agent data access.
• Retrieval-augmented agents accessing internal knowledge sources via vector search and document indexing infrastructure.
• Integrations with vendor agent platforms such as Microsoft Copilot Studio.

Define the schema contracts, governance policies, and validation approaches required for agentic systems to operate safely against regulated data.

Process discovery and improvement

Engage with stakeholders across the organization to identify process pain points, distinguish technical problems from organizational ones, and design appropriate interventions (whether software, workflow changes, or a combination). Treat process discovery as a deliverable in its own right.

Team leadership and technical direction

Lead and grow the team of AI solution engineers, whether internal staff or external contributors. Set technical standards and review practices, allocate effort across competing opportunities, and mentor engineers in the design discipline this role embodies. Hold the bar on quality, delivery, and judgment, and create the conditions for others to do their best work.

Evaluation, observability, and trust

Define how agent and solution quality is measured, both before release and in production. Build evaluation suites, monitoring, and feedback loops that surface drift, regressions, and failure modes early. Establish the guardrails, audit trails, and human-oversight patterns that make AI systems trustworthy enough to operate against regulated data.

Cross-system architecture and integration

Design and implement integration layers (including semantic layers, metadata services, and APIs) between NAP’s data platform, analyst-facing tooling, agent surfaces, and third-party systems. Define schema contracts, error semantics, and validation policies where systems cross team or trust boundaries. Author architectural documentation and maintain its alignment with the implemented systems over time.

Application and tool development

Build and maintain custom applications, internal tools, dashboards, and automations where vendor solutions are not a fit. Select technologies (including languages, frameworks, and modalities outside the core stack) based on what the problem requires.

Technical communication and decision documentation

Produce architecture explainers, decision records, and stakeholder-facing materials that serve engineers, analysts, and executives. Maintain documentation accuracy as systems evolve and reconcile prior specifications when they drift from current implementation.

What success looks like in the first year

In the first year, success in this role looks like:

• The unified metadata API and semantic layer are in production, serving agent consumers across clinical, commercial, and enterprise data.
• At least one AI agent is live for internal users, supported by an evaluation and monitoring framework that catches drift and regressions before they reach users.
• A prioritized map of high-friction processes exists, with the first interventions designed, built, and measured in production.
• Schema contracts, governance policies, and architectural documentation are established and adopted by the teams whose systems they govern.
• The solution-engineering team — internal staff and external contributors — is structured and operating against a clear, shared technical standard.

Required Qualifications

• Demonstrated track record of designing and delivering production software systems end-to-end, from problem framing through deployment and iteration.
• Hands-on experience building AI/LLM-based systems, including agents, retrieval-augmented generation (RAG) architectures with vector search and document indexing, prompt and context engineering, agent-tool protocols (such as MCP), and the operational concerns associated with deploying them.
• Strong software engineering fundamentals across multiple languages and stacks, with proficiency in Python and TypeScript / JavaScript, along with associated cloud and infrastructure-as-code tooling.
• Working knowledge of modern data platforms, including BigQuery or equivalent cloud warehouse, dbt, and semantic-layer concepts.
• Experience with cloud platforms (both GCP and Azure are currently in use at NAP), including service deployment, IAM, secrets management, and storage and search services.
• Demonstrated ability to lead process-discovery work with non-technical stakeholders, including scoping a problem, conducting stakeholder interviews, and producing a written framing that withstands review.
• Strong written and architectural communication skills, with a track record of producing documentation usable by varied audiences.
• Demonstrated executive-level judgment and a high degree of self-direction. The role does not operate from a defined task queue; the holder is expected to identify where their skills create the most leverage, make directional decisions within their domain, and apply their effort accordingly.

Preferred Qualifications

• Experience in regulated industries, including pharmaceuticals, life sciences, healthcare, or finance, and familiarity with the audit, validation, and access-control requirements they entail.
• Familiarity with clinical-trial data, CDISC standards (ADaM, SDTM), or comparable domain-standardized data models.
• Prior experience as a consultant or as a solo technical contributor within a small team. The role rewards generalists with depth in several areas more than it rewards narrow specialists.
• Exposure to one or more of the following: audio or media processing, BI platforms, mobile application development, knowledge-management systems. These are illustrative of the kind of breadth the role benefits from rather than strict requirements.

Salary and Benefits:

We offer a competitive base salary, annual bonus, and long-term incentives. In addition, we provide a comprehensive benefits package, including health insurance, dental and vision coverage, term life and disability coverage, and retirement plans.

NewAmsterdam Pharma is an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to sex, gender identity or expression, sexual orientation, marital status, race, color, national origin, ancestry, ethnicity, religion, age, veteran status, disability, genetic information, or any other basis protected by federal, state or local law.