AI Engineer

Key Responsibilities

Build Claude Skills for the Company

•
Design, write, and maintain Claude skills (SKILL.md files with YAML frontmatter and supporting scripts) that encode repeatable workflows our teams use every day.

•
Interview stakeholders across the business to identify tasks worth turning into skills — from editorial workflows to sales outreach to data pulls.

•
Follow skill authoring best practices: clear when-to-use descriptions, progressive disclosure, bundled reference files, and embedded scripts when needed.

•
Test, evaluate, and iterate on skills — measure whether they actually improve outputs versus a baseline prompt, and tune accordingly.

•
Maintain a central, documented skill library with versioning and clear usage guidance.

Build Programs with Claude Code

•
Use Claude Code as a daily collaborator to build internal tools, automations, scripts, and small applications.

•
Write AGENTS.md and CLAUDE.md files that teach agents our coding standards, repo structure, and team conventions.

•
Develop agentic workflows that combine Claude Code with MCP (Model Context Protocol) servers for Salesforce, Google Workspace, Slack, GitHub, and internal systems.

•
Apply the Plan / Execute / Review loop — decompose work into scoped tasks, run them through agents, review carefully, and ship.

•
Ship real code: pull requests, tests, documentation, deployment — working alongside senior engineers for mentorship and code review.

Enable AI Adoption Across the Company

•
Be the internal go-to person for "how do I get Claude to help me with this?"

•
Run training sessions, office hours, and hands-on workshops for non-technical teams on prompting, Claude Projects, skills, and agent use.

•
Write and maintain internal documentation: prompt libraries, skill directories, how-to guides, and best-practice playbooks.

•
Partner with team leads to identify high-value AI use cases in Editorial, Sales, Marketing, Events, Finance, and Operations — then prototype solutions.

•
Stay current on AI product releases (Claude, Claude Code, MCP ecosystem, browser agents, etc.) and translate what matters into actionable recommendations for the business.

Prototype and Evaluate AI Features

•
Build quick proofs-of-concept using the Anthropic API — Python, JavaScript, or whatever fits.

•
Design and run evaluations: what does "good" output look like, what percentage of the time does the workflow succeed, and what happens in the failure cases?

•
Think probabilistically. AI features are not deterministic, and part of your job is designing around that reality.

•
Integrate AI workflows with existing systems via MCP connectors, simple APIs, and light glue code.

Collaborate and Document

•
Write clearly — in Slack, in docs, in skill descriptions, in PR comments. Written communication is central to this role.

•
Demo your work frequently, gather feedback, and iterate.

•
Contribute to internal AI policy and responsible-use guidelines.

Our AI Stack

•
Core AI Platform: Claude (claude.ai, Claude Projects, Claude Code), Anthropic API

•
Skills & Agents: Claude skills (SKILL.md), subagents, MCP servers, Claude in Chrome

•
Integrations: Salesforce, Google Workspace (Drive, Gmail, Calendar), Slack, GitHub, n8n

•
Languages: Python and JavaScript/TypeScript (you should be comfortable reading and writing basic code in at least one)

•
Supporting Stack: PHP/Laravel, AWS, Redshift/MySQL/PostgreSQL, Docker — you are not expected to know these on day one, but exposure will come quickly

•
Dev Environment: macOS, VS Code, Git/GitHub, Claude Code CLI

What are we looking for?

Required

•
Genuine, demonstrable enthusiasm for AI. Tell us what you have built, broken, or experimented with — side projects, automations, Claude Projects, custom GPTs, hobbyist scripts, anything.

•
Strong written communication. You can explain a technical concept in plain English and craft clear, specific prompts.

•
Comfort with basic programming. You can read and write simple Python or JavaScript, use the command line, and work with Git.

•
Structured thinking. You can break a messy problem into steps and describe each one clearly — this is the core skill behind both prompting and skill authoring.

•
Willingness to work in public. You ask questions in Slack, demo rough drafts, and share what you learn.

•
Curiosity and self-direction. When you hit something you do not know, your first instinct is to go figure it out — often with Claude as your study partner.

•
Attention to detail. Small changes in prompts, skill descriptions, and instructions matter a lot.

Nice to Have (none of these are required)

•
A portfolio of AI experiments — custom GPTs, Claude skills or Projects, agent workflows, Discord/Slack bots, browser automations, etc.

•
Experience using Claude Code, Cursor, GitHub Copilot, or another AI coding assistant on a real project.

•
Familiarity with APIs and JSON — making a REST call, parsing a response.

•
Exposure to SQL. We live in Redshift and will teach you, but any prior SQL helps.

•
Experience writing technical documentation, tutorials, or how-to content.

•
A background in teaching, training, customer support, or writing — anyone who has explained technical ideas to non-technical people.

•
Awareness of the MCP ecosystem or agent frameworks.

•
A degree in a technical field is welcome but not required — we hire based on skills, not credentials.

What's in it for you?

• Competitive compensation structure!

• Medical, Dental and Vision Insurance - effective the 1st of the month following your start date.

• Short and Long Term Disability Insurance - this includes maternity and paternity leaves as well.

• 401K Profit Sharing Plan

• Flexible Spending Account

• Health Savings Account

• Dependent Care Account

• Unlimited Vacation Days

• 7 days paid sick leave

• 9 paid Holidays

• Referral Bonus Program

• Pet Friendly Offices

• You’ll get to work with incredibly smart, passionate, driven, ambitious, kind, caring and mindful people and will rarely experience an overload of policies, bureaucracy or toxicity (the latter, we do not tolerate).