AI product & program
40 postings read
AI-product postings open with a company mission paragraph anchored in a specific AI narrative (agentic, AI-native, foundation models, copilots), then pivot to a role description built almost entirely around ownership verbs (own, drive, define, ship) applied to roadmaps, platforms, or 'zero-to-one' problem spaces. They lean heavily on ambiguity as a selling point ('no playbook exists,' 'influence without authority') and treat hands-on use of AI tools/agents by the candidate as a qualification in itself, distinct from domain expertise.
- Company/mission preamble ~110w
- Team/context intro ~90w
- Role summary ~100w
- Responsibilities / What you'll do ~160w
- Requirements (must-have) ~90w
Research
40 postings read
AI-research postings open with a mission-driven company paragraph, then pivot immediately into a dense, bullet-heavy description of a specific technical problem (model efficiency, sim-to-real transfer, causal inference, agentic pipelines) rather than generic duties, and requirements are split into explicit tiered lists distinguishing baseline technical fluency from research pedigree. They lean heavily on stacking named methods, frameworks, and venues (PyTorch, RLHF, NeurIPS, TensorRT) as shorthand for credibility rather than describing skills in plain language.
- Company/Mission blurb ~90w
- Team/Lab description ~90w
- Role overview ~100w
- Responsibilities/What you'll do ~150w
- Minimum/Required qualifications ~90w
Safety, policy & governance
40 postings read
These postings cluster around technical AI-safety, trust-and-safety, and AI-governance roles at frontier labs, platform companies, and enterprise-governance vendors, mixing deep ML/research language with compliance/regulatory language depending on whether the role sits in research, engineering, or policy/legal. Nearly all open with an inflated mission statement about the company before ever describing the job, then shift into dense, unpunctuated bullet lists of responsibilities and qualifications with heavy jargon (RLHF, red-teaming, taxonomies, dual-use, frontier risk, monitorability) and almost never mention compensation.
- Company/mission preamble ~90w
- Team description ~80w
- About the Role / Your Impact ~100w
- Responsibilities / What You'll Do ~150w
- Qualifications (Minimum/Basic and Preferred/Bonus) ~100w
Data & analytics
40 postings read
These postings open with a paragraph of corporate self-description (mission, market position, scale metrics) before pivoting to a data/analytics/data-science role defined mainly through a long bulleted list of responsibilities framed as verbs (own, drive, define, partner, establish). They consistently foreground experimentation, metrics/KPIs, dashboards, pipelines, and cross-functional partnership, with AI/agentic work now a near-universal responsibility even in traditional analytics or data-science titles.
- Company/Mission Overview ~110w
- Team/Org Context ~80w
- About the Role / Opportunity ~100w
- What You'll Do / Key Responsibilities ~180w
- What You'll Bring / Qualifications ~100w
Machine learning & applied science
40 postings read
Machine-learning postings open with a company mission paragraph (usually tying the company's product to some frontier of AI) before naming the team and its ownership area, then list responsibilities as long unbulleted-feeling bullet strings heavy on verbs like 'own,' 'drive,' and 'architect.' Requirements are split into a terse must-have list (years of experience, specific frameworks/languages, degree) followed by an optional 'bonus/preferred' list that often name-drops publications, conferences, or specific tools.
- Company/mission overview ~110w
- Team description ~90w
- Role framing / About the role ~80w
- What you'll do / responsibilities ~150w
- Minimum/required qualifications ~80w
MLOps & infrastructure
40 postings read
These postings describe infrastructure and platform engineering roles for AI/ML systems—GPU fleets, inference serving, training pipelines, and the automation/observability layers around them—written by companies racing to scale compute for frontier models. They open with a boast about scale or mission (chip size, GPU count, user count, valuation) before narrowing into a specific team's ownership of one layer of the stack.
- Company/mission intro ~90w
- About the team ~80w
- About the role ~70w
- Responsibilities ~150w
- Requirements/Qualifications ~100w
