Magnit
Website:
magnitglobal.com
Job details:
Who We Are
Magnit is the future of work. Serving hundreds of the world’s most recognizable brands for the past 30+ years, Magnit offers the industry’s first holistic platform for the modern workforce. Magnit's integrated workforce management (IWM) platform supported by data, software, intelligence, and best-in-class services team is key to our clients’ success. It can adapt quickly to regional or industry economic shifts, and provides the speed, scale, flexibility, transparency, and expertise required to meet an organization’s contingent workforce management, talent strategy and broader organization goals. At Magnit, you’ll work with passionate colleagues who collaborate and deliver meaningful results that positively transform the largest companies around the globe.
Principal AI Engineer
About The Role
We’re hiring a
Principal AI Engineer to lead how we design, build, and productize generative and agentic AI, both inside our SaaS platform and across our products. This is a hands-on technical leadership role: you’ll own the architecture for production AI agents, set the standards teams follow to take AI from prototype to production, and mentor engineers along the way.
The hard problems here aren’t model demos. They’re reliable, safe, observable agent systems running in production: orchestration, memory, guardrails, evaluation, and human-in-the-loop control. You’ll be the person we trust to define what good looks like for AI engineering, and to make sure we build it responsibly.
What You’ll Do
- Own the architecture for generative and agentic AI across our SaaS platform, and stay hands-on in the most complex, critical components.
- Design and ship multi-agent systems that reason, plan, and execute, including the agentic loops, control flow, and failure isolation that keep them reliable in production.
- Build agent harnesses, reusable agent skills/tools, and MCP servers as the standard tool and integration layer between models and our systems, APIs, and data.
- Architect RAG systems (including GraphRAG) that ground outputs in trusted data, and event-driven pipelines that trigger and coordinate agents at scale.
- Productize agents end-to-end with guardrails, confidence-score-based routing, human-in-the-loop state machines, evaluation harnesses, tracing, and circuit breakers.
- Embed responsible-AI practices (privacy, bias mitigation, and auditability) directly into the reasoning and execution loops.
- Set architecture and code-quality standards, run design and code reviews, and mentor engineers; steer teams toward the right technologies and away from hype-driven choices.
- Make pragmatic build/buy decisions across frameworks, model providers, and orchestration patterns, balancing reliability, latency, security, and cost.
- Build with AI coding tools and AI-assisted development to raise team velocity and quality, with disciplined review gates.
What You’ll Bring
- 10+ years in software engineering, with a strong track record shipping production SaaS and large-scale, cloud-native distributed systems (AWS, GCP, or Azure).
- 5+ years in technical leadership (Staff / Principal / Lead / Architect) driving cross-team decisions and mentoring engineers.
- Hands-on experience building GenAI applications: LLMs, prompt and context engineering, embeddings, vector databases, and RAG.
- Proven experience designing and shipping agentic AI systems: multi-agent orchestration, tool/function calling, agentic loops, memory, and autonomous execution, both embedded in a product and standalone.
- Depth in AI productization: guardrails, confidence-based logic, human-in-the-loop control, evaluation, observability, and reliability of agents in production.
- Experience building MCP servers (or equivalent tool layers), reusable agent skills, and event-driven architectures.
- Strong Python plus one of Go, Java, or TypeScript; fluency with modern AI and agent frameworks (e.g., LangGraph, LangChain, CrewAI, AutoGen).
- Practical experience building products with AI coding tools.
- Working knowledge of responsible AI: privacy, bias, and auditability.
Nice to Have
- Building, fine-tuning, and deploying Small Language Models (SLMs) for routing, extraction, classification, and tool-calling, including on-device/edge and privacy-sensitive deployments.
- MLOps / LLMOps experience and observability/tracing stacks (e.g., OpenTelemetry).
- Open-source contributions, patents, or publications in AI.
- MS or PhD in Computer Science, Machine Learning, or a related field (or equivalent experience).
What Magnit Will Offer You
At Magnit, you’ll be joining an innovative, high-growth environment and can quickly make an impact to help transform the largest companies in the world. You will work with passionate colleagues who collaborate and deliver. Magnit offers all employees the opportunity for growth and development, and we want individuals to fulfill their potential and blaze their own trails!
Magnit will offer you a competitive PTO and benefits package, including medical, dental, and vision coverage, retirement planning, as well as discounts and perks for tickets, travel, merchandise and more! Magnit encourages employees to participate in giving back, and we will match employee contributions to favorite charities and support corporate volunteering hours to make a difference in your community!
If this role isn’t for you
Stay in touch, we will let you know when we have new positions on the team.
To see a complete list of our open career opportunities please visit.
https://magnitglobal.com/us/en/company/careers.html
To do our best work we need different viewpoints. Therefore, we celebrate diversity and embrace inclusion.
As an equal opportunity employer, we are dedicated to building a team that represents a variety of backgrounds, perspectives, and skills. We strive to ensure that we maintain a positive and enriching work environment for all.
By applying to this role, you consent to Magnit safely storing and managing your personal data. Please read this link to learn more.
https://magnitglobal.com/us/en/privacy-notice.html
Click on Apply to know more.