AI Innovation Engineer

Overview

As AI Innovation Engineer at 829, you'll work across the organization—partnering with VPs, directors, sales leadership, and subject matter experts—to drive operational excellence through AI-native tools and workflows. Your work directly impacts the team's ability to deliver cutting‑edge services to clients, building technology with real impact internally and externally.

This is not a traditional AI strategy role. You’ll write Python, build agentic workflows, ship MCP integrations, and use coding assistants (Claude Code, Cursor, Codex) as a daily workflow. We’re hiring a builder/coach who ships production AI work and earns adoption from subject matter experts.

6 Months:

• You’ve established a draft playbook around agentic workflow development inclusive of testing, implementation, and adoption.
• You’ve successfully crafted 3‑5 workflows that have met adoption criteria.
• You are able to quantify the success of 1‑2 workflows through quantitative and qualitative data.
• You have crafted a system to catalogue workflows, skills, and AI tools that enables shorter time‑to‑value and reduces sprawl.
• Your peers, colleagues and 829’s leadership team regularly seek your counsel in relation to AI service enablement.

12 Months:

• You have a pressure‑tested playbook that details the agentic workflow implementation process and ongoing management.
• You have established a framework for evaluating the performance of workflows.
• You have integrated multiple workflows into a custom front‑end in coordination with 829’s Head of Apollo and engineering resources.
• You have supported the growth of the innovation function as a center of excellence and / or an extension of 829’s data platform, Apollo.

Workflow & Agentic Engineering (60%)

• Playbook: Develop and manage a playbook that includes discovery, pilot, improvement, implementation, adoption and measurement phases to execute work.
• Pilots & Proof of Concept: Lead pilots across multiple disciplines to test approaches and technology, using results to inform production workflows and gain buy‑in.
• Integrations & Data Pipelines: Develop and maintain integrations between AI tools and core business systems and advertising platforms, including API builds, validation frameworks, and data pipelines.
• SME
  
  Collaboration:
  
  Collaborate closely with subject matter experts to ensure solutions and workflows deliver value, enable team members to produce their best work, are usable, and have access to appropriate data.
• Leadership
  
  Collaboration:
  
  Collaborate with senior leaders to ensure that workflows accomplish desired strategic and quality objectives.
• Adoption & Enablement: Ensure that workflows and systems are consistently improved and adopted through systems and cadences that check outputs, measure progress toward goals, and revisit training and adoption mechanisms.

AI Stack Curation / Governance / Documentation (15%)

• AI Stack Documentation: Build and maintain 829’s AI stack—tool inventory, vendor strategy, build‑vs‑buy posture, integration map, and shared‑context architecture.
• Skill Library: Develop reusable patterns, agentic workflows, and MCP integrations that provide strategic value and compound across workflows.
• Build vs. Buy: Consult on agency decisions to build workflows and tools vs. buy point solutions to achieve objectives.

General Enablement & Service‑Model Translation (25%)

• Internal Enablement: Run lightweight internal training, office hours, and troubleshooting clinics to upskill team members on AI‑augmented work.
• Service‑Model Translation: Advise/consult on revised service‑model artifacts: scope‑of‑work content, pricing implications, policies, and marketing collateral.
• Compliance & AI Policy: Advise and actively contribute to AI policies designed to encourage and enable use, establish SOPs, and mitigate risk to clients, staff, and the business.

Must Haves:

• 5+ years of professional experience with at least 2 years of exposure in the marketing / agency space.
• Engineering foundation strong enough to evaluate tools, build agentic workflows, ship MCP integrations, and use coding assistants (Claude Code, Cursor, Codex) as daily…