Principal Applied AI Engineer

About the role

Principal Applied AI Engineer

Fully Remote role, anywhere in the US

$235K–$290K + Token + Equity

Shields Group Search is partnering with a fast-growing consumer AI company built around principles of privacy, free speech, and user sovereignty.

The company is building a private and permissive AI ecosystem where millions of individuals and AI agents can gather, interact, and access sophisticated AI resources.

Joining the team means working alongside unorthodox builders who believe in moving quickly and delivering a beautiful mass-market consumer product that doesn’t spy on people or censor their ideas and questions. If you’re energized by big ideas, entrepreneurial spirit, individual empowerment, and the opportunity to help shape a fast-growing company in one of the world’s hottest industries from the ground up, this is an opportunity worth exploring.

Compensation: $235K–$290K base + company token + equity

Location: Remote - US based

Why They’re Hiring

The company is growing fast. Revenue is accelerating, the team is expanding, and the demands on every department — from engineering to marketing to operations to finance — are scaling with it.

They believe the next leap in company efficiency won’t come from hiring their way out of every bottleneck. It will come from giving every team the tools to build and deploy AI agents that multiply what they can accomplish.

They’re hiring a Principal Applied AI Engineer to build the internal agent platform that makes this possible across the entire company.

This is not a role scoped to engineering tooling alone. The departments that need the most help right now — marketing, support, communications, operations, and finance — are outside of engineering. You’ll build the platform they use to help themselves.

This is a solo individual contributor role to start, reporting directly to a Co-Founder. As the company grows and the platform matures, this role has a clear path to building and leading a team.

You’ll be the person who decides how the company builds its internal agent platform: whether that means wiring up commercial solutions — as long as they run on the company’s inference infrastructure — building orchestration frameworks from the ground up, or some combination.

The space is novel. There may not be commercial solutions yet. That’s fine. You’ll help figure it out.

What You’ll Do

Build the internal agent platform.

Design and ship the foundational infrastructure that teams across the company use to create, deploy, and manage their own AI agents. This includes agent templates, shared tool libraries, retrieval systems, orchestration patterns, eval pipelines, and the abstractions that let non-engineers assemble useful agents without writing code.

The goal: departments growing their output via agents they built on your platform, not solely via headcount.

Decide the build-vs-orchestrate strategy.

Evaluate commercial agent platforms, frameworks, and tools. Where existing solutions work and can run on the company’s inference infrastructure, use them. Where they don’t exist or don’t meet the company’s needs, build from the ground up. You own this decision space.

Dogfood the company’s own platform.

The internal agent platform you build should run on the company’s own inference API and models. The company is its own first customer.

This isn’t a branding exercise; it’s a product development advantage. The agents your colleagues deploy internally will surface edge cases, performance needs, and feature gaps that make the company’s platform better for external developers too.

Solve the data privacy problem.

This is the biggest constraint. The company’s brand is built on privacy and user sovereignty.

Internal agents must operate without leaking company data to third-party model providers. You’ll architect the platform so that agents run on the company’s own inference, data stays within its infrastructure, and access controls prevent cross-department data leakage.

This is not an afterthought; it is a core design requirement from day one.

Make it usable by non-engineers.

The platform’s success depends on adoption. Marketing managers, support leads, and finance analysts should be able to use your platform to build agents that solve their own problems.

That means clean abstractions, sensible defaults, good documentation, and a UX layer that hides complexity without removing power.

You’re building a product, and your internal colleagues are your users.

Partner with every team to understand their needs.

Work directly with marketing to understand their content and research workflows. Work with support to understand triage and response patterns. Work with operations on workflow automation. Work with finance on data extraction and reporting.

You’ll need to understand each team’s pain points well enough to design platform capabilities that let them build agents that actually help.

Evaluate, iterate, and prove ROI.

Measure the impact of the platform across the company. Close the loop between production traces, evaluations, and platform design. Optimize for quality, latency, and cost.

Show the company, in concrete terms, how agents built on your platform are multiplying human output.

Evangelize and educate.

Help non-engineering teams understand what’s possible with the platform. Run workshops, write internal docs, and build proof-of-concept demonstrations that make the abstract concrete.

The adoption challenge is as important as the technical one.