ResumeAdapter
How the Taleo parser works
Updated 2026-09-03

Taleo reads your resume once.
Then it fills in the form for you.

Why this matters

Oracle documents exactly what the Taleo parser extracts, and calls the parser itself a third party partner service, delivered as-is. That single sentence explains more about Taleo than any advice article does. This page walks the two stages Oracle describes, shows the field list and what is missing from it, and separates Taleo from the Oracle product that actually has the AI.

Field-by-field readMissing keywordsRewrite plan
By the numbers
Parse
2 stages
Extract, then seed the form
Parser
Third party
Delivered as-is, per Oracle
AI on your resume
None
The stars ship on Oracle Recruiting Cloud
Fully parsed
English
Plus English U.K. only

Quick answer

How does the Taleo ATS read your resume?

In two documented stages. First, Taleo converts your file to text and extracts a fixed list of fields: name, address, city, zip, phone, email and place of residence; education level, institution, program and dates; employer, job function, responsibility and employment dates. Oracle says parsing has no impact on formatting and that it can detect text in the header and footer, and it describes the parser as a third party partner service delivered as-is. Second, those extracted fields seed the application form that follows the upload, and Oracle specifies that all mapped fields are populated whether or not they are displayed to you. No AI is involved at any point: no machine-learning matching or scoring feature exists anywhere in Taleo Enterprise documentation .

The two-stage parse

Extraction, then form seeding.
Two stages, two failures.

Oracle Taleo has no branded parser and no model behind it. What Oracle documents is a third party extraction service and the block order that decides what happens to its output. Each stage does one job, and each fails in one specific way you can plan around.

  1. 01
    Stage

    Extraction: your file becomes a fixed set of fields

    Taleo converts your upload to text and pulls out a documented, finite list of fields. Personal: first, middle and last name, address, city, zip or postal code, home phone, email, place of residence, and the pasted resume itself. Education: education level, institution, program, start date, graduation date. Work experience: employer, job function, responsibility, start date, end date. Oracle also states that parsing has no impact on formatting, meaning bold, italics and bullets carry no signal, and that the parser can detect text in the header and footer. Oracle describes the whole component as a third party partner service delivered as-is.

    The failure it causes

    Look at what is missing from that list. There is no achievement field, no bullet field, no metric field. A bullet saying you cut month-end close by 40% has nowhere structured to land: it survives only as searchable text inside the pasted resume and your attachment. The other failure is upstream of all of it, because an image-format upload produces no text to extract and a language outside the supported set extracts almost nothing.

  2. 02
    Stage

    Form seeding: the extracted fields fill in the application

    The extracted fields do not replace the application form, they populate it. Oracle requires the Resume Upload block to sit alone on its page before the Attachments, Candidate Personal Information, Education, Work Experience and Resume blocks, precisely so the upload can seed what comes after it. Candidates are always shown the upload block when the education or experience blocks of a submission are empty, and the same mechanism carries LinkedIn and Indeed profile imports. Before the record is created, Taleo runs a duplicate check against existing candidate files.

    The failure it causes

    This is the failure almost nobody warns you about. Oracle documents that all fields mapped by the import service are populated whether or not those fields are displayed in the flow, and that where a field is not configured in the flow, the candidate will not have the opportunity to correct or edit that specific information. A mis-parsed employer, title or date can be written to your record, searchable by the recruiter, with no screen where you could fix it. If you abandon the application part way, Oracle says valid imported data is still saved to your record, with the application flagged incomplete.

Notice which rule is absent. Oracle documents nothing about multi-column layouts, tables, text boxes or graphics in Taleo, so every confident claim you have read about columns here cites nothing. For the checklist that is grounded in the documentation, see the Taleo resume format guide.

There is no semantic layer here. Oracle states that keyword search is considered to provide exact matches, with search terms joined by AND unless the recruiter switches to OR or NOT. A trailing asterisk works as a wildcard, so comput* finds computer, computing, computers and computation. There is also an optional related-terms mode, which Oracle describes as retrieving words with the same first six letters: administration also finds administrative and administrator.

Six letters is the entire flexibility budget. It will not carry project manager to project management, and it will not carry an acronym to its expansion. What is searched is broader than most people assume: with the optimal setting, Oracle lists personal information fields, the pasted cover letter and resume, career objectives, education and work experience fields, and all file attachments in DOC, DOCX, TXT, RTF, PDF, HTM, HTML, WPD and ODT. The older legacy setting narrows the attachments to the three most recent per candidate.

So the lever is vocabulary, not volume. Mirror the posting's own spelling for tools, certifications, titles and methods, and say each one once inside the role where you used it. Oracle documents no relevancy score and no frequency weighting anywhere in Taleo.

The AI that is not here

Three Oracle AI features,
none of them on Taleo.

No AI or machine-learning matching or scoring feature appears in any Oracle Taleo Enterprise documentation. The star ratings, the intelligent matching and the hiring agents that people routinely call “Taleo's AI” are all documented for Oracle Recruiting Cloud, a separate Oracle product on a different domain. Here is each one, and what its absence means when you apply on a taleo.net address.

Top Candidate Suggestions

Oracle Recruiting Cloud, not Taleo
What it is

This is the feature people picture when they talk about an AI match score at Oracle. In Oracle Recruiting Cloud, a recruiter viewing a requisition can see suggested candidates rated from zero to three stars, with the rating broken out across Profile, Skills, Experience and Education. It is documented for Oracle Fusion Recruiting. It appears nowhere in the Taleo Enterprise documentation.

What it means for your resume

The most widely read explanation of how Taleo ranks you on the internet is describing these stars. It is describing the wrong product. If your apply URL ends in taleo.net, there is no star rating on your application, no skills inference, and no model reading your resume. What ranks you is the prescreening criteria the employer set on the requisition.

The IRC_AI_INTELLIGENT_MATCHING profile option

Opt-in, and off on government pods
What it is

Intelligent matching in Oracle Recruiting Cloud is not on by default. It is gated behind a profile option an administrator has to enable, named IRC_AI_INTELLIGENT_MATCHING, and Oracle publishes a hard prerequisite alongside it: the Recruiting environment must not be on a government pod.

What it means for your resume

Two things follow. Even inside Oracle Recruiting Cloud, AI matching is a switch an employer may never have flipped, so you cannot assume it is running. And the exclusion lands exactly where Taleo is most concentrated, because state and municipal government is the strongest remaining Taleo footprint. Public-sector hiring is the least likely place to meet an Oracle matching model, on either product.

AI agents and Oracle Career Coach

Announced 15 October 2025
What it is

Oracle announced agentic AI for hiring on 15 October 2025, including an Oracle Career Coach, and its Fusion Cloud Recruiting 26B release documents AI agents that assemble candidate pools. Every capability ships in Oracle Fusion Cloud Recruiting. Taleo is not named in the announcement, and Taleo Enterprise documentation stops at 25B with no 25C, 25D or 26A release set.

What it means for your resume

This is the clearest public signal of where the two products are going. Oracle is putting its AI investment into Oracle Recruiting Cloud while Taleo, closed to new customers as of Oracle's 1 February 2026 price list, keeps running the deterministic rules it always ran. Applying through Taleo in 2026 means applying to a system that reads text and checks answers, and nothing more than that.

Sources: Oracle Fusion Cloud Recruiting documentation, Top Candidate Suggestions and the IRC_AI_INTELLIGENT_MATCHING profile option with its government-pod prerequisite; Oracle Fusion Cloud Recruiting 26B What's New, AI Agent candidate pools; Oracle, agentic AI for hiring announcement (2025-10-15); Oracle Taleo Cloud Service Global Price List (2026-02-01). Accessed 2026-09-03.

Oracle's Fusion Cloud Recruiting 26B release documents AI agents that assemble candidate pools for recruiters, following the agentic-AI hiring announcement Oracle made on 15 October 2025 that introduced an Oracle Career Coach. Every one of those capabilities ships in Oracle Fusion Cloud Recruiting. Taleo is not named. Taleo Enterprise documentation stops at 25B with no 25C, 25D or 26A release set, and the Taleo Business Edition 25C What's New has a single revision-history row recording that the document was created, listing no features at all. That is the split an applicant needs to understand: Oracle's AI work goes into Oracle Recruiting Cloud, while Taleo keeps extracting text and checking prescreening answers.

What the parser keys on

Five things the parse
actually depends on.

Everything above reduces to five levers, and every one of them traces to a sentence Oracle published. Four decide whether the parse reads you at all. The fifth decides your ranking, and it has nothing to do with the document.

  1. 01
    Depends on

    Content that maps to a documented field

    Why and where to fix it

    The extraction list is fixed: names, address, contact, education level, institution, program, employer, job function, responsibility and dates. Anything outside it, including every achievement and every number, is stored only as searchable text. Write the mappable parts plainly and put the rest in language a recruiter would search for. Fix it in the Taleo resume format guide.

  2. 02
    Depends on

    English, or a much thinner parse

    Why and where to fix it

    Oracle lists resume parsing as fully supported only for English and English U.K., limited for Dutch, French, German, Italian, Portuguese (Brazil) and Spanish, and unsupported for roughly twenty-five other languages. The default career-section parsing scope is English only, so a resume in an unsupported language uploads cleanly and extracts nothing. Fix it in the format guide.

  3. 03
    Depends on

    A real text layer, never an image

    Why and where to fix it

    Taleo Business Edition added PNG and JPEG resume uploads, and Oracle's own release note says the resume will not be parsed when it is in an image format. The application still appears to succeed. Use DOCX, PDF, RTF, TXT, HTML or ODT and confirm you can select the text with your cursor. Fix it in the format guide.

  4. 04
    Depends on

    The posting's exact words, spelled its way

    Why and where to fix it

    Oracle states that keyword search is considered to provide exact matches, with terms joined by AND by default. The only flexibility is a trailing wildcard and an optional related-terms mode that matches words sharing the first six letters, so administration also finds administrator. That will not bridge project manager to project management, and repetition buys nothing. Fix it in the format guide.

  5. 05
    Depends on

    The prescreening answers, which outrank the document

    Why and where to fix it

    Ranking does not come from the parse. The employer marks criteria Required, meaning minimum requirements, or Asset, meaning strongly preferred. Meeting all Required plus some Asset criteria earns the ACE Candidate flag recruiters filter on, and a disqualification question can instantly exit you from the process. Fix it in the Taleo ATS guide.

The fastest way to see which of these five your resume already passes, and which it fails, is to read it the way a parser does. .

See the parse, not the layout

See the record a parser builds from your resume.

The scanner reads your resume the way an extraction step does, then shows the fields that come back empty, the exact terms the posting uses that your resume never says, and the bullets that need a number. Free, no signup to see the score.

FAQ

How Taleo reads your resume FAQ

The questions candidates ask when they want to know what Taleo actually does with the file they upload. Every answer is grounded in Oracle's own Taleo documentation, and all eight are byte-identical to the FAQPage JSON-LD, because AI engines that extract HTML and AI engines that extract JSON-LD should not see different text.

How does the Taleo ATS read your resume?

In two documented stages. First it extracts text into a fixed field set: name, address, city, zip, phone, email, place of residence and the pasted resume; education level, institution, program, start and graduation dates; employer, job function, responsibility and employment dates. Second, those extracted fields seed the application form, which Oracle requires to sit after the resume upload block. Oracle describes the parser itself as a third party partner service delivered as-is, and says parsing has no impact on formatting, so bold, italics and bullets carry no signal either way.

What fields does Taleo actually extract from a resume?

Oracle publishes the complete list. Personal: first name, middle name, last name, address, city, zip or postal code, home phone number, email address, place of residence, and the pasted resume. Education: education level, institution, program, start date, graduation date. Work experience: employer, job function, responsibility, start date, end date. There is no achievement field, no bullet field and no metric field anywhere in that list, which is why your strongest quantified results survive only as searchable text in the attachment and the pasted resume rather than as structured data a recruiter can filter on.

Does Taleo use AI to match candidates to jobs?

No. No AI or machine-learning matching or scoring feature appears in any Oracle Taleo Enterprise documentation. Taleo matches on structured fields the employer marks Required or Desired, such as preferred jobs, preferred locations, competencies, job level, schedule and education level. The zero-to-three-star candidate suggestions rated on Profile, Skills, Experience and Education belong to Oracle Recruiting Cloud, a different Oracle product on a different domain, and even there they sit behind an administrator-enabled profile option that Oracle documents as unavailable on government pods.

Can you correct a field Taleo filled in wrong from your resume?

Only if the employer chose to display that field. Oracle states that all fields mapped by the profile import service are populated whether or not those fields are displayed in the submission flow, and that where a field is not configured in the flow, the candidate will not have the opportunity to correct or edit that specific information. So a mis-parsed employer, job function or date can be stored on your candidate record, searchable by the recruiter, with no screen where you could ever fix it. Read back and correct every field you can actually see before you advance.

Does the Taleo parser count keyword frequency?

No frequency weighting is documented anywhere in Taleo. What Oracle does document is the recruiter-side keyword search: it is considered to provide exact matches, terms are joined by AND unless the recruiter changes it, a trailing asterisk acts as a wildcard, and an optional related-terms mode retrieves words sharing the same first six letters, so administration also finds administrative and administrator. Six letters will not bridge project manager to project management. Saying a term once, in the posting's own spelling, is what the search rewards; saying it five times is not.

Will Taleo parse a resume that is not written in English?

Usually not well. Oracle's resume parsing language table lists only English and English U.K. as fully supported. Dutch, French for Canada and France, German, Italian, Portuguese for Brazil and worldwide Spanish are listed as limited. Roughly twenty-five further languages, including Japanese, Korean, Chinese, Polish, Swedish and Turkish, are not supported at all. The career section setting that controls this, Resume Parsing Languages and Scope, ships with a default of English only, so a resume in an unsupported language uploads without error and extracts nothing.

Does Taleo parse a resume uploaded as a photo or screenshot?

No. Taleo Business Edition added PNG and JPEG resume uploads, and Oracle's release note for that change says plainly that the resume will not be parsed when it is in an image format. The upload appears to succeed and the application goes through, but the record it creates carries no extracted fields and no searchable text. That is worse than a rejection, because nothing tells you it happened. Use a supported text-based format instead and confirm you can select the text with your cursor before you submit.

Is the Taleo parser the same as the Oracle Recruiting Cloud parser?

They are different products and should not be treated as one. Taleo career sections live on a company.taleo.net host, or on a code.tbe.taleo.net host for Taleo Business Edition. Oracle Recruiting Cloud lives on an oraclecloud.com host with hcmUI/CandidateExperience in the path. Everything on this page is documented for Taleo. The star-rated suggestions, the intelligent matching profile option and the AI agents announced through 2025 are documented for Oracle Recruiting Cloud. Click through to Apply or Sign In and read the address bar, because many employers white-label Taleo behind their own careers domain.

Continue

Go deeper on the parts
that need their own page.

Pillar

The Oracle Taleo ATS resume guide

The hub: eight rules Oracle actually documents, why ACE prescreening decides your ranking, who still runs Taleo, and a free scan against a Taleo posting.

Read the pillar
Spoke

Taleo resume format, do-this and avoid-this

The documented checklist: supported file types, the 1 MB attachment limit, the English-only parsing default, and the fields to read back before you submit.

Read the spoke
Spoke

Companies that use Taleo

The employer list verified by live apply URL, the big names commonly mislabelled as Taleo when they run Oracle Recruiting Cloud, and how to check any employer yourself.

Read the spoke
Resume example

A real resume scored before and after

See a single-column, parser-safe resume scored against a job description, with the keyword and metric fixes that moved the ATS score, annotated line by line.

See the example
Before you apply through Taleo

Run your resume against
the actual Taleo job description.

Get the parser-readable check, the exact posting terms your resume never says, the bullets that need a number, and a rewrite plan. Free to scan; no signup to see the score.

Your resume stays yoursResume and analysis data may be saved to your account. Resume data is never sold.Taleo-readable check