DESCRIPTION FORMATTING RULES (CRITICAL - FOLLOW EXACTLY)

Your job is to output clean, readable Markdown while PRESERVING ALL ORIGINAL CONTENT AND STRUCTURE EXACTLY.

ABSOLUTE RULES - NEVER VIOLATE:

NEVER add, remove, or modify any text/wording from the original
NEVER paraphrase or rewrite sentences - keep exact wording
NEVER remove sections, paragraphs, disclaimers, or any content
PRESERVE all links exactly as they appear
PRESERVE the natural flow and spacing of the original document

MARKDOWN FORMATTING:

for main section headings (bold standalone titles like "The Role", "You Will", "Requirements")
for sub-section headings
NEVER use # (H1) - the job title is already the page H1
Use "-" bullet lists for ANY list of items
CRITICAL: If content is clearly a list (requirements, responsibilities, benefits), format as bullet list
Bold for inline emphasis, key terms, skills - preserve existing bold from original
Italic for emphasis or disclaimers - preserve existing italic from original
text for ALL hyperlinks - preserve every link

SPACING RULES:

Preserve the natural paragraph breaks from the original content when they exist
IF the original lacks proper spacing (e.g., everything is one giant paragraph), ADD logical breaks:
- Blank line before each section heading
- Blank line between distinct topics/sections
- Blank line before and after bullet lists
NO blank lines between bullet points in the same list
NO multiple consecutive blank lines (max 1 blank line between elements)
The output should be scannable and readable, even if the input wasn't

OUTPUT QUALITY:

Clean, scannable markdown that matches the original document's flow
Professional formatting that readers can quickly parse
All content intact - nothing added, nothing removed

TITLE FORMATTING RULES (CRITICAL)

The "title" field MUST use the EXACT format: "[Job Title] - [Company Name]"

Always use a HYPHEN (" - "), NEVER "at"
Example: "Senior ML Engineer - OpenAI" (CORRECT)
Example: "Senior ML Engineer at OpenAI" (WRONG — never use "at" in the title field)
Clean up the raw title: fix capitalization, remove company name if already embedded, remove location info
Keep the job title portion concise but don't lose important qualifiers (e.g., "Senior", "Lead", "Staff")
The "metaTitle" field uses a DIFFERENT format ("[Job Title] at [Company]") — do NOT confuse the two

SALARY DISPLAY FORMAT (IMPORTANT)

Format salary display as compact "k" notation:

$187,000 - $220,000 → "$187k - $220k"
$100,000 - $120,000 USD → "$100k - $120k"
€80,000 - €100,000 → "€80k - €100k"
Always use "k" for thousands, remove ".0" decimals
Include currency symbol at the start, not the end

WORK MODE INFERENCE RULES

Determine workMode using these rules IN ORDER:

If explicitly stated ("remote", "hybrid", "on-site", "in-office") in title/description → use that
If location contains "Remote" → REMOTE
If multiple countries/continents listed → REMOTE (high confidence)
If location is specific city only (e.g., "London", "New York", "Tokyo") with no remote mention → ONSITE
If "flexible" or "work from home options" mentioned → HYBRID
If unclear after all checks → null

EXPERIENCE LEVEL EXTRACTION RULES

Extract experienceLevel using these rules:

From title keywords (highest priority):
- "Intern/Internship" → ENTRY
- "Junior/Jr/Associate" → JUNIOR
- "Mid/Mid-level" → MID
- "Senior/Sr/Staff/Principal" → SENIOR
- "Lead/Tech Lead/Team Lead" → LEAD
- "Director" → DIRECTOR
- "Head of/Head" → HEAD
- "VP/Vice President/Chief/C-level/CTO/CEO/CFO" → C_LEVEL
From years requirement in description:
- 0-1 years → ENTRY
- 1-2 years → JUNIOR
- 3-4 years → MID
- 5+ years → SENIOR (unless title indicates higher level)
If no indicators → null

LOCATION RULES (CRITICAL - keep compact)

Goal: Locations must be SHORT and COMPACT. Max 5 comma-separated values total.

REMOTE JOBS:
- If open to all/most regions → just "Remote"
- If remote but limited to specific regions → use the CANONICAL REMOTE TOKENS below in parentheses
- If remote limited to 2 regions → "Remote (US, Europe)" (still 1 badge)
- NEVER list every individual country/city for remote jobs
CANONICAL REMOTE TOKENS (use these EXACT values inside parentheses):
- United States only → "Remote (US)"
- US + Canada / North America → "Remote (NA)"
- Europe / EU / EMEA / DACH / Nordics → "Remote (Europe)"
- Asia / Asia Pacific / APAC / SEA / ANZ → "Remote (APAC)"
- Latin America / South America / LATAM → "Remote (LATAM)"
- Middle East / MENA / GCC → "Remote (MENA)"
- Africa / Sub-Saharan Africa → "Remote (Africa)"
- UK only → "Remote (UK)"
- Specific country → "Remote (Singapore)", "Remote (Germany)", etc.
ONSITE/HYBRID with specific cities:
- For each city, ALSO include the country as a separate entry for discoverability
- Example: "Dubai" → "Dubai, United Arab Emirates"
- Example: "London, Berlin" → "London, United Kingdom, Berlin, Germany"
- If 4+ cities in same country → use country name: "United States" and at most 2 key cities
- NEVER include both city AND state AND country for the same place (e.g., "New York" not "New York, NY, United States")
- But DO include city AND country as separate comma entries: "New York, United States"
US-SPECIFIC:
- NEVER list states separately (no "CA, NY, TX, FL...")
- Many US cities → "United States" or top 2 cities + "United States"
- Remote in US → "Remote (US)"
MAX 5 comma-separated values in the locations field

TAG RULES (CRITICAL - content-derived skills only)

Goal: Tags must be SPECIFIC, ACTIONABLE skills/technologies/tools that appear IN THE JOB DESCRIPTION. They surface in "In-Demand Skills" and must help candidates filter by real requirements — not generic industry words.

FORBIDDEN TAGS (NEVER use these — they are useless on an AI job board):

"ai", "artificial intelligence", "machine learning", "tech", "product management", "business development", "marketing", "engineering", "design", "operations", "sales", "partnerships", "product marketing"
Any tag that merely restates the job category or industry (we already know it's AI/ML)
Any tag not explicitly mentioned or clearly implied by the job description text

REQUIRED: Tags must be EVIDENCE-BASED. Only include a tag if the description (or title) explicitly mentions that skill, tool, protocol, or practice. If the post doesn't mention specific technologies or methods, use 1–2 tags maximum or leave tags minimal.

GOOD TAGS (prefer these when present in the content):

AI/ML domains: "nlp", "computer vision", "reinforcement learning", "generative ai", "llm", "robotics", "autonomous systems"
Languages & runtimes: "python", "rust", "typescript", "go", "c++", "java", "cuda"
Frameworks & tools: "pytorch", "tensorflow", "jax", "hugging face", "langchain", "ray", "mlflow", "wandb"
Tools & practices: "figma", "wireframes", "product specs", "user research", "a/b testing", "sql", "react", "node"
Disciplines (only when clearly a focus): "ui/ux", "mlops", "data engineering", "compliance", "devrel", "deep learning"
Concrete deliverables/activities: "prds", "roadmaps", "model evaluation", "dashboards", "fine-tuning"

RULES:

Do NOT repeat words already in the job title (e.g. "Product Manager" → no tag "product management").
Prefer lowercase; well-known abbreviations are fine: "nlp", "ui/ux", "llm".
Max 3 tags, prefer 2. If nothing specific is stated, 0–1 tags is fine — do not fill with generic terms.
When in doubt, omit. Fewer accurate tags are better than generic ones.

RELEVANCY SCORING (for feed ranking)

Score how relevant this job is to an AI/ML job board audience (0-100).
This score controls feed ranking — it directly determines which jobs appear
first when a company has many openings. Be precise and consistent.

95-100: Core AI/ML research or engineering that BUILDS models
(ML Engineer, AI Researcher, NLP Scientist, Computer Vision, RL, LLM, MLOps)
80-94: Software infrastructure that directly SUPPORTS AI systems
(GPU/CUDA Engineer, ML Platform, Distributed Training Infra, AI DevOps)
65-79: Technical roles at AI companies (Software Engineer, Backend, Security, SRE)
50-64: Product, data, and design roles tied to AI products
(AI Product Manager, UX Researcher on AI tools, Data Analyst)
30-49: Business and go-to-market roles (Sales, Marketing, Finance, Legal, Recruiting)
10-29: Administrative and support roles (Office Manager, Coordinator, Compliance)
0-9: Non-technical facility/manual roles (Electrician, Custodian, HVAC, Security Guard)

Key heuristic: Would someone browsing an AI job board be excited to find this listing?
A "Senior ML Engineer" at Anthropic = 98. A "Facilities Coordinator" at Anthropic = 5.

AI RELEVANCE CLASSIFICATION (CRITICAL — controls whether job is listed)

Classify whether this job involves AI/ML/LLM/agentic technologies.
This is a strict binary gate for companies that aren't pure AI companies.

Set "isAIRelevant": true if ANY of these apply:

The role is fundamentally about AI/ML (researcher, ML engineer, NLP scientist, etc.)
The job description requires building, training, deploying, or evaluating AI/ML models
The role involves significant use of LLMs, RAG, agentic frameworks, or AI APIs
The role directly supports AI systems (ML platform, AI infrastructure, GPU engineering)
The role is an AI product manager, AI designer, or AI-focused data scientist
The description explicitly lists AI/ML tools (PyTorch, TensorFlow, LangChain, etc.)

Set "isAIRelevant": false if:

The role is generic software engineering with no AI/ML mention in the description
The role is business/operations/admin with no AI/ML component
The role is at a company that does AI but THIS specific position doesn't involve AI
Example: "Recruiter" at Stripe — no AI involvement even though Stripe uses AI
Example: "Backend Engineer" at Datadog — builds dashboards, no ML mentioned
Example: "Office Manager" at any company — never AI-relevant

Be strict: a job at an AI-adjacent company is NOT automatically AI-relevant.
The role or description must explicitly involve AI/ML/LLM technologies.

EXTRACTION RULES ===

Extract data ONLY if clearly stated
If not found, return null (not empty string, not "Competitive")
Use the structured data from job board when available (more reliable than parsing description)
For categoryId: ALWAYS select best matching category based on title and description

AVAILABLE CATEGORIES (id:name):
1:Business Development, 2:Data, 3:Design, 4:Engineering, 5:Finance, 6:Legal, 7:Marketing, 8:Operations, 9:People Ops, 10:Product Management, 11:Research, 12:Sales, 14:Support

EXPERIENCE LEVELS: ENTRY, JUNIOR, MID, SENIOR, LEAD, DIRECTOR, HEAD, C_LEVEL

JOB LISTING TO ANALYZE ===

Title: AI Research Engineer – Datadog AI Research (DAIR)
Company: Datadog
Location: New York, New York, USA
Description:
As a Research Engineer on our team, you will partner with Research Scientists to turn research ideas into working systems, building the data, tooling, and infrastructure that enable rapid iteration, trustworthy evaluation, and a smooth path from prototype to production.

Building on our track record of AI-powered solutions (e.g., Bits AI, Bits Evolve, and our time series foundation model), Datadog AI Research tackles high-risk, high-reward problems grounded in real-world challenges in cloud observability and security.

We are focused on two research areas:

World Models for Observability -- Training multimodal foundation models that learn the joint dynamics of distributed systems across metrics, traces, logs, topology, and events. These models power advanced forecasting, anomaly detection, root cause analysis, counterfactual simulation ("what if? "), and provide a learned planning backbone for our autonomous agents.
Trained Agents for Observability-- Post-training models to operate autonomously across Datadog's domain. SRE incident response is our first target, with a clear path to code repair, security response, and infrastructure optimization. We build the simulation environments, RL training loops, and evaluation infrastructure needed to train agents that match or surpass frontier models at a fraction of the cost.

What You'll Do:

Build and operate multimodal data pipelines, training and evaluation infrastructure, benchmarks, and internal tooling
Implement models, run experiments at scale, and profile for reliability, performance, and cost
Build simulation environments and replay infrastructure for agent training and evaluation
Orchestrate distributed training and distributed RL with Ray, including scheduling, scaling, and failure recovery
Establish rigorous automated benchmarks and regression tests for world model predictions, agent performance, and simulation fidelity
Collaborate with Research Scientists, Product, and Engineering to integrate capabilities into Datadog's products and to harden prototypes into reliable services
Contribute to research publications at top-tier conferences (e.g., NeurIPS, ICLR, ICML), and produce high-quality code, documentation, and open-source artifacts

Who You Are:

You have depth in distributed computing, RL Infra, and ML systems for training and inference at scale; experience with Ray, Slurm, or similar frameworks is a plus
You are proficient in Python, familiar with a systems language (e.g., Rust, C++, or Go), and comfortable with modern cloud and data infrastructure
You have practical experience implementing and operating ML training and inference systems (e.g., PyTorch or JAX), including containerization, orchestration, and GPU acceleration
You have practical experience with large-scale model training and fine-tuning, including frameworks like Megatron-LM, DeepSpeed, SkyRL, VeRL, or TorchTitan, and techniques such as SFT, RLVR, RLHF, and efficient inference (quantization, speculative decoding)
You can explain design and performance trade-offs clearly to both technical and non-technical audiences
You have experience supporting or contributing to research publications

Bonus Points (any of the following):

You have strong software engineering skills with experience in domains such as observability, SRE, or security
You have experience bridging research prototypes and real-world product applications, especially with large foundation models, world models, or RL-trained agents
You have a passion for pushing the boundaries of AI with a focus on customer impact and scalable deployment
You have hands-on experience with GPU programming and optimization, including CUDA
You have experience writing production data pipelines and applications
You have experience building simulation or sandbox environments for agent training

Datadog values people from all walks of life. We understand not everyone will meet all the above qualifications on day one. That's okay. If you’re passionate about technology and want to grow your skills, we encourage you to apply.

Benefits and Growth:

Competitive global benefits
New hire stock equity (RSUs) and employee stock purchase plan (ESPP)
Opportunity to collaborate closely with colleagues across the Datadog offices in New York City and Paris
Opportunity to attend and present at conferences and meetups
Intra-departmental mentor and buddy program for in-house networking
An inclusive company culture, ability to join our Community Guilds (Datadog employee resource groups)

Benefits and Growth listed above may vary based on the country of your employment and the nature of your employment with Datadog.

About Datadog:

Datadog (NASDAQ: DDOG) is a global SaaS business, delivering a rare combination of growth and profitability. We are on a mission to break down silos and solve complexity in the cloud age by enabling digital transformation, cloud migration, and infrastructure monitoring of our customers’ entire technology stacks. Built by engineers, for engineers, Datadog is used by organizations of all sizes across a wide range of industries. Together, we champion professional development, diversity of thought, innovation, and work excellence to empower continuous growth. Join the pack and become part of a collaborative, pragmatic, and thoughtful people-first community where we solve tough problems, take smart risks, and celebrate one another. Learn more about #DatadogLife on Instagram, LinkedIn, and Datadog Learning Center.

Equal Opportunity at Datadog:

Datadog is an Affirmative Action and Equal Opportunity Employer and is proud to offer equal employment opportunity to everyone regardless of race, color, ancestry, religion, sex, national origin, sexual orientation, age, citizenship, marital status, disability, gender identity, veteran status, and more. We also consider qualified applicants regardless of criminal histories, consistent with legal requirements. Here are our Candidate Legal Notices for your reference.

Your Privacy:

Any information you submit to Datadog as part of your application will be processed in accordance with Datadog’s Applicant and Candidate Privacy Notice.

#LI-Hybrid

Datadog offers a competitive salary and equity package, and may include variable compensation. Actual compensation is based on factors such as the candidate's skills, qualifications, and experience. In addition, Datadog offers a wide range of best in class, comprehensive and inclusive employee benefits for this role including healthcare, dental, parental planning, and mental health benefits, a 401(k) plan and match, paid time off, fitness reimbursements, and a discounted employee stock purchase plan.

The reasonably estimated yearly salary for this role at Datadog is:

$140,000—$400,000 USD

About Datadog:

Equal Opportunity at Datadog:

Datadog is proud to offer equal employment opportunity to everyone regardless of race, color, ancestry, religion, sex, national origin, sexual orientation, age, citizenship, marital status, disability, gender identity, veteran status, and other characteristics protected by law. We also consider qualified applicants regardless of criminal histories, consistent with legal requirements. Here are our Candidate Legal Notices for your reference.

Datadog endeavors to make our Careers Page accessible to all users. If you would like to contact us regarding the accessibility of our website or need assistance completing the application process, please complete this form. This form is for accommodation requests only and cannot be used to inquire about the status of applications.

Privacy and AI Guidelines:

Any information you submit to Datadog as part of your application will be processed in accordance with Datadog’s Applicant and Candidate Privacy Notice. For information on our AI policy, please visit Interviewing at Datadog AI Guidelines.

Datadog

AI Research Engineer – Datadog AI Research (DAIR) - Datadog

DESCRIPTION FORMATTING RULES (CRITICAL - FOLLOW EXACTLY)

for main section headings (bold standalone titles like "The Role", "You Will", "Requirements")

for sub-section headings