Going deeper on the top 3 modern:
1. Greenhouse

How Greenhouse uses AI

• Greenhouse has built-in AI features for things like resume parsing, anonymization (blurring identifying candidate details), helping with job description writing, suggesting keywords to recruiters, and scorecard-based evaluation. Jobscan+3my.greenhouse.com+3Greenhouse Support+3

• They explicitly mention that during “application screening” the AI helps suggest search terms and filters for the recruiting team. my.greenhouse.com+1

• On the resume anonymization side: the feature uses machine-learning / deep-learning to identify and conceal certain fields (name, email, etc) so evaluation can be more equitable. Greenhouse Support

• Also, Greenhouse integrates third-party AI screening/“resume shortlisting” tools (e.g., integrations like “AI Screened” that connect into Greenhouse) which further process incoming applications and score them. integrations.greenhouse.com+1
  

How you should prepare for Greenhouse

Given how their AI behaves:

• Keywords matter: Make sure your resume matches the role description in terms of key skills, job title variations, domain language. The AI suggests filters/search terms, so if you include the same terms recruiters are likely expecting, you’re more likely to match.

• Structure and clarity: Since parsing and anonymization happen, ensure your resume is clearly formatted (standard headings like “Experience”, “Education”, “Skills”) so the system can correctly extract data. Avoid exotic/uncommon formatting.

• Match role requirements: If the job description lists “B2B SaaS product manager, 5+ yrs, user research, agile”, make sure those phrases or equivalent are clearly in your experience. The AI and recruiters will be looking for fit.

• Focus on measurable impact: Given that scoring and screening may depend on evidence of “meets / does not meet” criteria, provide concrete results (e.g., “Grew user base 60% in 12 months”, “Reduced churn from 7% to 4%”) rather than vague statements.

• Use standard file types: Consistently use common resume formats (PDF or DOCX) unless otherwise specified. Ensures parsing works well.

• Avoid hidden/unreadable text: Avoid embedding text in images, odd fonts, or columns that might confuse the parsing engine.

• Tailor per job: Even if you submit to many roles, adjust your resume for each to align with the specific role language. Because AI screening makes this feasible and likely.
  

2. Ashby

How Ashby uses AI

• Ashby explicitly describes an “AI-Assisted Application Review” that lets a hiring team define objective criteria for a job and then the AI analyzes resumes to indicate whether each criterion is met or not. Ashby+2Ashby+2

• They emphasize transparency & responsible AI: The AI generates citations (i.e., shows “why” it thinks the applicant meets a criterion) and human-review is still required. Ashby+1
  

• They also mention integrations for AI-powered candidate screening, AI for outreach, but for resumes specifically the “meets/does not meet” criteria is key. Ashby
  
  

How you should prepare for Ashby

• Understand the required criteria: Because the hiring team can set specific “criteria” (e.g., “3+ years product marketing in enterprise SaaS”, “experience working remote across time zones”, “Proficient in Jira & Confluence”), you want your resume to explicitly show those criteria. If you don’t mention them, you risk being filtered out or flagged as “does not meet”.

• Be explicit and structured: Use bullet points where you mention specific tools, environments, metrics, timelines. The AI will try to map your text to the criteria.

• Avoid ambiguity: For example, instead of just “led product launch”, write “led B2B SaaS product launch to 5,000 customers in 6 months, achieving X% growth”. This gives clearer evidence the criterion is met.

• Match terminology: If the job description uses “Go-to-Market strategy”, “buyer persona research”, “growth marketing”, use those phrases if they apply to your background. Because the AI filter may key off those words.

• Keep your keywords in experience summary too: It helps if your top summary or heading mentions the core skill set so the AI doesn't miss it buried too deep.

• Clean and standard layout: As with any ATS, avoid unusual layout or graphic resumes that might confuse scanning.

3. Lever

How Lever uses AI

• Lever’s system includes a resume parser: it extracts name, organization, contact details, work history, etc from uploaded resume files and populates candidate profile fields. help.lever.co

• Lever also supports “AI Companion / AI agents” or integrations that enable automated candidate screening, matching candidate profiles to job requirements, personalized outreach, and data-management. getguru.com+1

• For example, when integrated with e.g. “AI Screened”, Lever can automatically run screening evaluations, surface top candidates, etc. help.lever.co
  

How you should prepare for Lever

• Ensure your resume parsing works: Since Lever will pull details from your resume (name, contact, work history, dates), make sure those sections are plainly labelled (“Work Experience”, “Company”, “Dates”, etc.). If the parser fails, your profile may look incomplete.

• Prominently state your key role details: Because information is extracted to populate fields, make your company name, role title, dates, and main responsibilities easily identifiable. Avoid creative naming of sections that hide these details.

• Include quantifiable results: Although Lever’s AI may focus more on extraction and matching, having clear, numeric achievements helps you stand out when screeners review.

• Use relevant keywords: The AI agents that match candidates to requirements will likely look for keywords around the role. Make sure you include key skills/tools/languages that are required in the job description.

• Use a clean structure: Avoid tables, graphics, or multi-column formats. A standard vertical resume layout improves parsing reliability.

• Tailor for the role: Similar to Greenhouse and Ashby, make your resume role-specific. Use language from the job description to align with what the AI might match.
  
  

Key common preparation tips across all three

• Always tailor your resume to the job description — match language, list required tools/skills, showcase measurable impact.

• Use a simple, readable format (one-column, standard fonts, headings) so parsing & scanning AI/ATS systems don’t mis-read your resume.

• Include keywords from the job description — especially skills, tools, seniority level, industry domain — since screening filters often match those.

• Provide quantitative evidence (e.g., “increased revenue by 25%”, “managed team of 8”, “reduced churn from X% to Y%”) rather than vague statements.

• Use standard headings (Experience, Education, Skills) so ATS/AI can correctly identify and extract content.

• Avoid embedding important content in images, sidebars, or flashy layouts — those often break parsing.

• Submit correct file format (PDF or DOCX as allowed) and follow any specified file naming or upload instructions.

• Be honest — AI tools often flag inconsistencies or omissions, so mis-leading or false information can cause issues.

If you know the company uses one of these ATS, you can optimise accordingly, but you won’t always know — so the standard best practices above cover all cases.