Applied AI Engineer
Bandit · Brooklyn, NY · 3 wk ago
On-siteOTHR$150k–$170k/yrFull-time
We need to output HTML fragment with sections: About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule. Only include sections that have real specific content. We must not repeat job title. Must not include Job ID etc. Must remove boilerplate. Must preserve all substantive info.
We need to parse the posting.
Content:
- Intro about Bandit Running (mission, vision). That's about the company, maybe "About the role"? Actually "About the role" could be description of role. The intro about Bandit Running is more about company. Could be placed under "About the role" or maybe we can have a section "About the role" describing the role and company. The instruction: use for section headings (e.g. About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule). So we can have those headings if content exists.
We need to extract:
About the role: description of role: "As the fastest-growing brand in the sport, we challenge conventional brand-building norms... We are seeking an Applied AI Engineer to help build the systems that change how work gets done at Bandit. You'll partner closely with our Head of Data & Technology to identify the largest opportunities across the business and turn them into production AI tools, agents, and automated decision loops. This is a hands-on builder role, not a research or advisory role. You'll work across models, APIs, data, application logic, and deployment to take projects from problem → prototype → production system. Early use cases will span growth, marketing, and operations, with significant freedom to shape what we build and how we build it."
That's substantive.
Responsibilities: The bullet list under "Requirements" actually lists responsibilities? The posting has "Requirements" then bullet list of responsibilities: Build production AI systems end-to-end..., Design agent architecture..., Automate growth and marketing workflows..., Build personalization and decisioning systems..., Help define Bandit's applied AI strategy...
These are responsibilities/duties. So we can put them under Responsibilities.
Requirements: The "What We're Looking For" bullet list: 4+ years of software engineering experience..., 1+ years deeply immersed in applied LLM engineering..., Hands-on experience building agents..., Strong across Python/TypeScript..., Strong product and business judgment..., Comfortable operating broadly...
These are qualifications/skills. Could be split into Requirements and Qualifications and Skills. But we need to decide.
The instruction: Use headings for About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule. We should include sections only if there is real specific content.
We have content for About the role, Responsibilities, Requirements (maybe the "What We're Looking For" list), Qualifications (maybe overlapping), Skills (maybe technical stack). Benefits: listed: Competitive salary + equity compensation, 401k match, Top tier medical insurance/benefits, Rocket-ship level momentum with a career-defining opportunity at-hand. Also mention culture values but that's maybe not benefit. We'll include Benefits.
Pay: Salary range: $150,000 - $170,000 + equity. That's specific.
Schedule: Not mentioned. No schedule info. So we should not include Schedule.
Qualifications vs Skills: Could separate: Qualifications: years of experience, LLM immersion, etc. Skills: Python/TypeScript, APIs, SQL/databases, cloud infra, deployment, product judgment, etc. But we need to avoid duplication. We can put the "What We're Looking For" list under Requirements (or Qualifications). Then maybe Skills under a separate heading for technical stack. Let's see.
The "What We're Looking For" includes: 4+ years software engineering experience, 1+ years applied LLM engineering, Hands-on experience building agents..., Strong across Python/TypeScript, APIs, SQL/databases, cloud infrastructure, and application deployment, Strong product and business judgment, Comfortable operating broadly...
We could treat the first three as Requirements (experience), the technical stack as Skills, and product judgment as Qualifications? But it's ambiguous.
Simplify: Use Requirements for the bullet list (the whole "What We're Looking For") and skip Qualifications and Skills unless we have distinct content. But we have technical stack that could be Skills. However we could incorporate into Requirements as part of bullet. The instruction: Prefer a few substantive sections over many thin ones. So maybe we just have About the role, Responsibilities, Requirements, Benefits, Pay. That's five sections. That seems fine.
We must not include empty sections.
Thus output:
for section headings (e.g. About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule). So we can have those headings if content exists.
We need to extract:
About the role: description of role: "As the fastest-growing brand in the sport, we challenge conventional brand-building norms... We are seeking an Applied AI Engineer to help build the systems that change how work gets done at Bandit. You'll partner closely with our Head of Data & Technology to identify the largest opportunities across the business and turn them into production AI tools, agents, and automated decision loops. This is a hands-on builder role, not a research or advisory role. You'll work across models, APIs, data, application logic, and deployment to take projects from problem → prototype → production system. Early use cases will span growth, marketing, and operations, with significant freedom to shape what we build and how we build it."
That's substantive.
Responsibilities: The bullet list under "Requirements" actually lists responsibilities? The posting has "Requirements" then bullet list of responsibilities: Build production AI systems end-to-end..., Design agent architecture..., Automate growth and marketing workflows..., Build personalization and decisioning systems..., Help define Bandit's applied AI strategy...
These are responsibilities/duties. So we can put them under Responsibilities.
Requirements: The "What We're Looking For" bullet list: 4+ years of software engineering experience..., 1+ years deeply immersed in applied LLM engineering..., Hands-on experience building agents..., Strong across Python/TypeScript..., Strong product and business judgment..., Comfortable operating broadly...
These are qualifications/skills. Could be split into Requirements and Qualifications and Skills. But we need to decide.
The instruction: Use headings for About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule. We should include sections only if there is real specific content.
We have content for About the role, Responsibilities, Requirements (maybe the "What We're Looking For" list), Qualifications (maybe overlapping), Skills (maybe technical stack). Benefits: listed: Competitive salary + equity compensation, 401k match, Top tier medical insurance/benefits, Rocket-ship level momentum with a career-defining opportunity at-hand. Also mention culture values but that's maybe not benefit. We'll include Benefits.
Pay: Salary range: $150,000 - $170,000 + equity. That's specific.
Schedule: Not mentioned. No schedule info. So we should not include Schedule.
Qualifications vs Skills: Could separate: Qualifications: years of experience, LLM immersion, etc. Skills: Python/TypeScript, APIs, SQL/databases, cloud infra, deployment, product judgment, etc. But we need to avoid duplication. We can put the "What We're Looking For" list under Requirements (or Qualifications). Then maybe Skills under a separate heading for technical stack. Let's see.
The "What We're Looking For" includes: 4+ years software engineering experience, 1+ years applied LLM engineering, Hands-on experience building agents..., Strong across Python/TypeScript, APIs, SQL/databases, cloud infrastructure, and application deployment, Strong product and business judgment, Comfortable operating broadly...
We could treat the first three as Requirements (experience), the technical stack as Skills, and product judgment as Qualifications? But it's ambiguous.
Simplify: Use Requirements for the bullet list (the whole "What We're Looking For") and skip Qualifications and Skills unless we have distinct content. But we have technical stack that could be Skills. However we could incorporate into Requirements as part of bullet. The instruction: Prefer a few substantive sections over many thin ones. So maybe we just have About the role, Responsibilities, Requirements, Benefits, Pay. That's five sections. That seems fine.
We must not include empty sections.
Thus output:
About the role
...
Responsibilities
- ... ...
Requirements
- ... ...
Benefits
- ... ...
Pay
$150,000 - $170,000 + equity
We must not include schedule. Now ensure we preserve all substantive information: The intro about Bandit Running mission etc. Should we include that? It's about the company, not role. But we could include in About the role paragraph. The About the role section should contain the role description and maybe company mission. The instruction: "Do NOT repeat the job title as a heading (the page already shows it)". It didn't say we cannot include company description. So we can include it in About the role. Thus About the role paragraph: Combine the initial paragraph about Bandit Running mission and the role description. Let's craft: About the role: Bandit Running is a pioneering technical performance brand committed to creating innovative, high-quality apparel & gear for goal-driven runners. Our mission to "Evolve Running" guides everything we do, always putting the needs of the global running community first. As the fastest-growing brand in the sport, we challenge conventional brand-building norms, prioritizing an extremely high bar for disruptive excellence across product, experiences & content. Our vision is to build an enduring, iconic brand that breaks down barriers & inspires greater passion & participation in the sport we love. We are seeking an Applied AI Engineer to help build the systems that change how work gets done at Bandit. You'll partner closely with our Head of Data & Technology to identify the largest opportunities across the business and turn them into production AI tools, agents, and automated decision loops. This is a hands-on builder role, not a research or advisory role. You'll work across models, APIs, data, application logic, and deployment to take projects from problem → prototype → production system. Early use cases will span growth, marketing, and operations, with significant freedom to shape what we build and how we build it. That's all substantive. Responsibilities: list from Requirements bullet points. We need to ensure we don't lose any detail. Responsibilities bullet points: - Build production AI systems end-to-end — agents, internal applications, workflows, and decision systems that connect models with Bandit's data and operational tools - Design agent architecture and infrastructure — including tool use, permissions, context management, memory, triggers, scheduling, observability, and human-in-the-loop controls - Automate growth and marketing workflows — starting with systems that evaluate performance, make recommendations, generate or iterate creative, and accelerate experimentation across paid media, CRM, and site merchandising - Build personalization and decisioning systems that turn customer, product, membership, and behavioral signals into relevant actions and experiences - Help define Bandit's applied AI strategy by identifying where AI can materially increase team capacity, improve decision quality, or shorten the time between idea, experiment, and learning Now Requirements: from "What We're Looking For" bullet list: - 4+ years of software engineering experience, with a track record of shipping production systems. Depth of traditional software engineering is helpful, but not required - 1+ years deeply immersed in applied LLM engineering—actively following, testing, and adopting the best techniques emerging from frontier model labs and the broader AI engineering community - Hands-on experience building agents and LLM applications using tool calling, structured outputs, context/retrieval systems, evaluations, and multi-step workflows - Strong across Python/TypeScript, APIs, SQL/databases, cloud infrastructure, and application deployment - Strong product and business judgment—you can take an ambiguous problem, determine what should actually be built, and ship quickly without sacrificing reliability where it matters - Comfortable operating broadly and independently across engineering, data, growth, creative, and operational problems We could also include "Success in This Role" maybe as part of Responsibilities? But it's more about success metrics. Could be included in About the role or Responsibilities? It's not a responsibility but description of success. Could be omitted? The instruction: preserve ALL substantive information. So we should include it somewhere. Perhaps we can add a paragraph after Responsibilities under a heading like "Success in this role"? But not in allowed headings. Could incorporate into About the role or Responsibilities. Probably best to add a short paragraph under About the role after role description: "Success in this role means building systems that improve how Bandit operates: codifying strong judgment, enabling better decisions, accelerating experimentation, and shortening the loop between learning and action." Thus About the role will have two paragraphs: company description + role description + success paragraph. Now Benefits: list: - Competitive salary + equity compensation - 401k match - Top tier medical insurance/benefits - Rocket-ship level momentum with a career-defining opportunity at-hand We could also mention culture values but that's not a benefit; but it's substantive. Could be included in About the role? Already we have mission/vision. Could add a sentence about culture: "At Bandit, we're building an incredible culture rooted in our three core values—Trust, Empathy, & Excellence. We believe that in order to build the best team, our baseline is an environment where everyone can bring their full & authentic selves to the table. We are committed to fair hiring practices where we hire team members for their potential & advocate for diversity, equity, & inclusion. We do not discriminate based on your race, color, religion, gender identity or expression, sexual orientation, national origin, age, military service eligibility, veteran status, marital status, disability, or any other protected class." This is substantive but maybe not required. The instruction: preserve ALL substantive information. So we should include it. Where? Could be in About the role after success paragraph. Or we could add a separate section "Culture" but not allowed headings. So we can add to About the role. Thus About the role will have multiple paragraphs. Now Pay: "$150,000 - $170,000 + equity" Now ensure we don't include any boilerplate like cookie/navigation/legal/EEO/equal-opportunity/scam-warning. The EEO statement is the non-discrimination paragraph; that is substantive? It's a legal statement but maybe considered boilerplate. The instruction says remove EEO/equal-opportunity/scam-warning boilerplate. So we should remove the non-discrimination paragraph. It is EEO boilerplate. So we should not include that. Good catch. Thus we should not include the non-discrimination paragraph. Also we should not include "We are committed to fair hiring practices where we hire team members for their potential & advocate for diversity, equity, & inclusion." That's also EEO. So remove. Thus About the role will include mission, vision, role description, success statement, maybe culture values (Trust, Empathy, Excellence) which is not EEO but core values. That's okay. Now check for any other boilerplate: "At Bandit, we're building an incredible culture rooted in our three core values—Trust, Empathy, & Excellence. We believe that in order to build the best team, our baseline is an environment where everyone can bring their full & authentic selves to the table." That's not EEO, it's culture. Keep. Now ensure we didn't include any duplicated lines. Now produce HTML. We must output only HTML fragment, no markdown fences. Let's craft: