Legal AI Product Tester / Legal QA Analyst

Please login or register as jobseeker to apply for this job.

TYPE OF WORK

Part Time

SALARY

4.50 per hour

HOURS PER WEEK

30

DATE UPDATED

May 29, 2026

JOB OVERVIEW

Legal AI Product Tester / Legal QA Analyst

Remote | Contract | Philippines-Based Preferred

About the Role

Irys is looking for lawyers or legally trained professionals to test our legal AI platform on a daily basis. This role is focused on stress-testing legal AI workflows across litigation, corporate, transactional, research, drafting, and document-review use cases.

You will use your own legal judgment, test sets, sample documents, pleadings, contracts, memos, and research tasks to evaluate whether the system produces accurate, useful, well-structured, and lawyer-reviewable work product.

This is not a passive QA role. We need someone who can think like a lawyer, identify what went wrong, explain why it matters, and structure feedback clearly for our product and AI teams.

Responsibilities

Test Irys legal AI workflows daily across different legal tasks.
Bring and create realistic test sets, including pleadings, contracts, memos, client emails, research questions, diligence materials, and mixed document bundles.
Evaluate outputs for legal accuracy, factual grounding, completeness, structure, tone, formatting, and usefulness.
Identify hallucinations, missed facts, weak reasoning, unsupported claims, bad drafting choices, citation problems, and confusing product behavior.
Compare outputs across different modes, prompts, models, and workflows.
Provide structured feedback with clear issue descriptions, severity levels, screenshots, source references, and recommended fixes.
Work closely with product, engineering, and AI teams to reproduce issues and improve the system.
Help define testing rubrics and repeatable evaluation sets for future benchmarking.

What You Should Be Testing For

Did the system understand the uploaded documents?
Did it extract the right facts and miss anything important?
Did it distinguish between fact, law, argument, risk, and client narrative?
Did it cite or reference sources correctly?
Did it draft in the right legal format?
Did it over-admit, overstate, hallucinate, or make unsupported conclusions?
Did it produce something a lawyer could actually review and use?
Did the workflow make sense from a user perspective?
Did any internal/debug language appear in the final output?
Did the product behave reliably across repeated tests?

Qualifications

Law degree or substantial legal training required.
Strong familiarity with U.S. legal documents, pleadings, motions, contracts, legal memos, and common law reasoning strongly preferred.
Experience with U.S., U.K., Australian, Singaporean, or other common law legal work is a plus.
Litigation and/or corporate/transactional experience preferred.
Strong legal writing and analytical skills.
Excellent attention to detail.
Comfortable testing software, documenting bugs, and giving precise feedback.
Ability to work independently and test consistently without needing constant direction.
Interest in legal AI, legal technology, and product development.

Ideal Candidate

The ideal candidate is a legally trained operator who can test like a lawyer and think like a product reviewer. You should be comfortable saying:

This answer is legally wrong.
This draft is useful but not filing-ready.
This missed a key fact.
This cited the wrong source.
This workflow is confusing.
This is a serious issue versus a minor polish issue.

Engagement

Remote contract role.
Part-time or full-time depending on fit and availability.
Daily testing expected.
Paid trial period may be required.
Must be able to provide structured written feedback in clear English.

To Apply

Please send:

Resume or LinkedIn profile.
Short summary of your legal background.
Jurisdictions or legal systems you are familiar with.
Examples of legal documents or workflows you would be comfortable testing.
Availability and expected hourly or monthly rate.
Any prior experience with AI tools, legal tech, QA, or product testing.

VIEW OTHER JOB POSTS FROM:
SHARE THIS POST
facebook linkedin
  BENCHMARKS  
Loading Time: Base Classes  0.0008
Controller Execution Time ( Jobseekers / Job )  0.0179
Total Execution Time  0.0196
  GET DATA  
No GET data exists
  MEMORY USAGE  
1,505,472 bytes
  POST DATA  
No POST data exists
  URI STRING  
jobseekers/job/Legal-AI-Product-Tester-Legal-QA-Analyst-1657506
  CLASS/METHOD  
jobseekers/job
  DATABASE:  onlinejobs (Jobseekers:$db)   QUERIES: 15 (0.0111 seconds)  (Hide)
0.0005   SELECT *
                                
FROM exrates
                                WHERE rate_name 
'USD-PHP' 
0.0010   SELECT *
FROM `employer_jobs`
WHERE `job_id` = 1657506
 LIMIT 1 
0.0005   SELECT *
FROM `employers`
WHERE `employer_id` = 939601
 LIMIT 1 
0.0010   SELECT COUNT(*) AS `numrows`
FROM `t_thread` `t`
LEFT JOIN `t_thread_misc` `miscON `t`.`id` = `misc`.`thread_id`
WHERE `t`.`job_id` = 1657506
AND `misc`.`idIS NULL 
0.0004   SELECT COUNT(*) AS `numrows`
FROM `t_thread` `t`
LEFT JOIN `t_thread_misc` `miscON `t`.`id` = `misc`.`thread_id`
WHERE `t`.`job_id` = 1657506
AND `misc`.`idIS NULL 
0.0006   SELECT e.business_namee.logoe.websitee.rebill_datee.date_added member_datehitsDATEDIFF('2026-08-23',ej.date_added) duration_daysDATEDIFF('2026-08-23',e.rebill_date) duration_rebillej.*, e.deactivate FROM employers eemployer_jobs ej WHERE e.employer_id ej.employer_id AND
                                   ((
e.user_level >= '500' AND ej.date_added <= e.rebill_date)
                                   OR 
e.employer_id '' OR (ej.date_approved <> '2000-01-01' and DATEDIFF('2026-08-23',ej.date_added) <= 14 ))
                                   AND 
e.deactivate != AND ej.deleted AND job_id '1657506' 
0.0009   SELECT *
FROM `employer_jobs_skills` `ejs`
LEFT JOIN `skills_categories` `scON `ejs`.`skill_id` = `sc`.`id`
WHERE `job_id` = 1657506 
0.0004   SELECT COUNT(*) AS `numrows`
FROM `t_thread` `t`
LEFT JOIN `t_thread_misc` `miscON `t`.`id` = `misc`.`thread_id`
WHERE `t`.`job_id` = 1657506
AND `misc`.`idIS NULL 
0.0021   UPDATE employer_jobs SET hit_counts '***May-29-2026=600***May-30-2026=63***May-31-2026=42***Jun-01-2026=47***Jun-02-2026=45***Jun-03-2026=27***Jun-04-2026=16***Jun-05-2026=22***Jun-06-2026=18***Jun-07-2026=16***Jun-08-2026=23***Jun-09-2026=14***Jun-10-2026=6***Jun-11-2026=13***Jun-12-2026=5***Jun-13-2026=8***Jun-14-2026=11***Jun-15-2026=18***Jun-16-2026=19***Jun-17-2026=15***Jun-18-2026=17***Jun-19-2026=13***Jun-20-2026=14***Jun-21-2026=7***Jun-22-2026=11***Jun-23-2026=18***Jun-24-2026=11***Jun-25-2026=10***Jun-26-2026=8***Jun-27-2026=5***Jun-28-2026=7***Jun-29-2026=13***Jun-30-2026=14***Jul-01-2026=14***Jul-02-2026=4***Jul-03-2026=4***Jul-04-2026=3***Jul-05-2026=9***Jul-06-2026=5***Jul-07-2026=11***Jul-08-2026=7***Jul-09-2026=8***Jul-10-2026=3***Jul-11-2026=3***Jul-12-2026=5***Jul-13-2026=13***Jul-14-2026=10***Jul-15-2026=7***Jul-16-2026=10***Jul-17-2026=7***Jul-18-2026=5***Jul-19-2026=6***Jul-20-2026=5***Jul-21-2026=8***Jul-22-2026=12***Jul-23-2026=3***Jul-24-2026=10***Jul-25-2026=8***Jul-26-2026=4***Jul-27-2026=5***Jul-28-2026=10***Jul-29-2026=5***Jul-30-2026=9***Jul-31-2026=6***Aug-01-2026=4***Aug-02-2026=6***Aug-03-2026=7***Aug-04-2026=7***Aug-05-2026=6***Aug-06-2026=6***Aug-07-2026=7***Aug-08-2026=11***Aug-09-2026=6***Aug-10-2026=3***Aug-11-2026=2***Aug-12-2026=5***Aug-13-2026=1***Aug-14-2026=4***Aug-15-2026=2***Aug-16-2026=5***Aug-17-2026=6***Aug-18-2026=7***Aug-19-2026=9***Aug-20-2026=7***Aug-21-2026=3***Aug-22-2026=9***Aug-23-2026=1' WHERE job_id'1657506'  
0.0008   UPDATE employer_jobs SET monthly_hits '***May-2026=705***Jun-2026=471***Jul-2026=219***Aug-2026=124' WHERE job_id'1657506'  
0.0015   SELECT date_sent FROM jobseeker_sent_emails WHERE jobseeker_id '' AND job_id '1657506' AND status LIKE 'sent%' ORDER BY id DESC  
0.0004   SELECT *
FROM `employer_jobs_skills` `ejs`
LEFT JOIN `skills_categories` `scON `ejs`.`skill_id` = `sc`.`id`
WHERE `job_id` = 1657506 
0.0004   SELECT COUNT(*) AS `numrows`
FROM `employer_jobs`
WHERE `employer_id` = '939601'
AND `date_added` >= '2022-06-08' 
0.0004   select from teasers 
0.0002   SELECT FROM skill_categories WHERE skill_cat_id='' 
  HTTP HEADERS  (Show)
  SESSION DATA  (Show)
  CONFIG VARIABLES  (Show)