Agentic Lead / Applied AI Prototyper

IANS•Boston, MA

7h•$144,000 - $180,000•Hybrid

About The Position

The Agentic Lead / Applied AI Prototyper is a builder role embedded in New Product Development with one clear mission: prototype agentic workflows and AI-generated output formats that are grounded in IANS data and IP, move fast, and generate tangible artifacts the organization can evaluate, test, and scale. This is not a strategy or advisory position. The Agentic Lead spends the majority of their time building—experimenting with Claude and other LLM frameworks, assembling multi-step agentic workflows, and producing concrete outputs (board-ready memos, diagnostics, risk scorecards, structured briefings) that demonstrate what AI-enabled IANS content can look like at its best. Prototypes that demonstrate validated impact and repeatability graduate into the product backlog through a structured handoff process—ensuring that the Studio’s best work becomes durable product capability rather than orphaned experiments.

Requirements

Hands-on experience building LLM-powered applications or agentic workflows in a production or near-production context—not just experimentation.
Proficiency in Python and comfort working with LLM APIs (Anthropic Claude, OpenAI, or equivalent) including prompt engineering, function/tool calling, and structured output design.
Experience with at least one agentic or RAG framework: LangChain, LlamaIndex, AutoGen, CrewAI, or similar.
Strong instincts for what makes a good AI-generated output—able to evaluate quality, consistency, and fitness-for-purpose without always needing external validation.
Ability to work in a fast, iterative environment with minimal process overhead—comfortable with ambiguity, able to self-direct, and oriented toward shipping over theorizing.
Excellent written communication skills; able to document prototypes, playbooks, and workflows clearly enough that others can reproduce and build on them.

Nice To Haves

Experience working with RAG architectures including vector databases (Pinecone, Weaviate, pgvector, or similar) and embedding pipelines.
Background in content-heavy or knowledge-intensive domains (research, advisory, legal, financial services) where AI output quality and accuracy are non-negotiable.
Familiarity with multi-agent architectures and orchestration patterns for complex, multi-step reasoning workflows.
Experience producing structured, templated AI outputs (reports, scorecards, briefings) for professional audiences.
Exposure to cybersecurity, information security, or technology risk content—or the ability to learn quickly in a domain-specific environment.
Understanding of the IANS faculty model and how practitioner-grade advisory content is created, validated, and delivered to clients.

Responsibilities

Agentic Workflow Prototyping
Design, build, and iterate on multi-step agentic workflows using Claude, LangChain, LlamaIndex, AutoGen, or comparable frameworks—grounded in IANS research content, faculty IP, and proprietary data assets.
Develop workflow modules for high-priority use cases: AI-assisted meeting prep, automated CRM enrichment, research synthesis, client engagement plan generation, and renewal risk scoring.
Build and test prompt architectures, retrieval-augmented generation (RAG) pipelines, and structured output templates that produce consistent, high-quality, on-brand IANS artifacts.
Maintain a rapid iteration cadence—prototype, test with internal stakeholders, discard or refine, and move on. Speed and learning velocity are the primary performance metrics here.
Output Format Development
Produce working examples of AI-generated IANS output formats: board-ready security briefings, CISO diagnostics, risk scorecards, issue-based memos, and structured Ask-an-Expert summaries.
Collaborate with faculty members to validate AI-generated artifacts for accuracy, nuance, and practical relevance—ensuring outputs meet the standard clients expect from IANS.
Document the prompt patterns, data inputs, and workflow logic behind each successful output format so they can be reproduced, taught, and eventually scaled by the Enablement function.
Studio Operations & Knowledge Transfer
Maintain an active, documented library of prototypes, agentic playbooks, and output templates—organized for eventual handoff to the AI Enablement Lead for broader deployment.
Participate in structured handoff processes when Studio prototypes are ready for scaling—providing technical documentation, success criteria, and integration notes for the Ask IANS engineering team.
Share learnings across the Studio, Ask IANS, and Enablement functions—contributing to a shared understanding of what works, what doesn’t, and what’s next.
Engage directly with the CPO and NPD leadership to align Studio output priorities with organizational strategy.
Technical Exploration & Tooling
Continuously evaluate emerging LLM capabilities, agentic frameworks, and AI tooling relevant to IANS’s use cases—bringing informed recommendations to the team rather than defaulting to the familiar.
Assess and integrate new model capabilities (extended context, function calling, structured outputs, multi-agent coordination) as they become available.
Identify infrastructure or data layer requirements that would unlock higher-quality agentic outputs and collaborate with Ask IANS engineering to address them.