AI Solution Consultant / AI Architect

Armis; New York (NY); On-site; USD 120,000 - 180,000; Full time

Job summary

Armis, a security company in New York, is building an AI Center of Excellence. This role focuses on designing and delivering AI-native solutions—agents, assistants, and reusable skills—that advance Armis's security mission and scale across teams.

You will train frontier models, compose capable AI systems, and bring rigor to prove they work in production. Strong Python, agent frameworks, and cloud integration are essential.

Qualifications

• An AI expert in multi-agent solutions.
• Proven experience shipping real AI solutions (agents, assistants, copilots) that were used.
• Hands-on familiarity with modern agent frameworks and developer tooling (SDKs from major providers).
• Strong evaluation discipline with metrics to ensure quality, safety, and reliability.

Responsibilities

• Design, build, and productionize AI models, agents, assistants, and reusable skills.
• Turn ideas into MVPs and partner with engineering to harden them into production-ready systems.
• Define success metrics and perform rigorous evaluations of each solution.

Skills

AI multi-agent solutions

Production AI systems

Agent frameworks & tooling

Python

API & cloud integration

Education

Bachelor's or Master's in Computer Science/Engineering or related field

Tools

Docker

Kubernetes

AWS

Azure

GCP

Job description

Armis is seeking talented, motivated AI builders to join our Strategic Initiatives team as part of a new AI Center of Excellence.

Reporting to the VP of Strategic Initiatives, this team designs and ships AI-native solutions — agents, assistants, and reusable skills — that advance Armis's security mission and multiply the impact of teams across the company.

This is a builder role. We're looking for engineers fluent in modern AI application development — agent frameworks, harnesses, tool use, prompting, retrieval, and evaluation — not machine learning engineers. You will be expected to train foundation models / frontier models . You will compose capable, production-grade AI systems on top of frontier models, get them into people's hands, and bring the rigor to prove they actually work.

What you'll do

• Design, build, and bring to production ready: models, AI agents, assistants, and reusable skills that solve real problems for Armis and its customers.
• Turn ideas into working MVPs fast, then partner with engineering to harden them into reliable, production-ready systems.
• Build evaluation into everything you ship — define what "good" means for each solution, measure it, and hold the bar. Due diligence through rigorous evals is non-negotiable, not an afterthought.
• Compose agentic workflows that are secure by construction, using secure harnesses along with appropriate isolation, guardrails, and controls.
• Select and integrate the right frameworks, tools, and models for each job, and turn what works into reusable patterns and best practices for the COE.
• Also be a part of the COE Consult team to help across the organization.
• Document what you build and how it should be used, operated, and evaluated.

Qualifications

• An AI Expert in Multi-agentic solutions.
• Demonstrated experience building real AI solutions — agents, assistants, copilots, or skills — that shipped and got used, not just prototyped.
• Deep, hands-on familiarity with modern agent frameworks and developer tooling (agent SDKs such as the Top 3 providers SDKs (Claude, OpenA and Google), thorough understanding and use of SKILL.md skills standard, plus the judgment to know which to reach for when.
• Strong evaluation discipline: building test sets, defining metrics, and measuring quality, safety, and reliability before and after deployment.
• Proficiency in Python and comfort integrating APIs, tools, and cloud services (AWS, Azure, or GCP) with containerization (Docker, Kubernetes).
• Bachelor's or Master's in Computer Science, Engineering, or a related field, or equivalent professional experience.

Preferred

Experience in security, or building AI for security use cases