ContentQuo Evaluate MT

Manage your human evaluations of Machine Translation quality effortlessly at any scale

Perform quick human assessment with Adequacy-Fluency

Deeply analyze MT output with MQM error annotation

Run post-editing “lab tests” for Edit Distance

Human evaluation is valuable indeed but it is nor fast or cheap if you outsource it to trained professionals, which is what we do at CPSL. In this sense, moving from spreadsheets to ContentQuo has been a big leap for us.

All our professional evaluators prefer ContentQuo over spreadsheets, and our customers benefit from the speed with which we can send our proposals and scalability helps us keep evaluation costs within budget.

Lucía Guerrero
Machine Translation Specialist, CPSL  
process

Where does it fit in my MT workflow?

Most companies implementing Machine Translation run human quality evaluation after they are satisfied with automatic quality metric values — e.g. on shortlisted engines when selecting an engine mix, on newly retrained engines when assessing training results, or on post-edited translations on a regular basis to identify recurring errors that could be fixed with training. ContentQuo Evaluate MT makes it easy, fast, and efficient.

talk to an expert
Run
automatic
metrics
shortlist
BEST
engines
train
BEST  
engines
Select or
retrain
engines
BENEFITS

Why use specialized MT software instead of Excel spreadsheets?

90
%
less overhead
Your team has better things to do than doing boring manual work like  copy&paste into Excel and emails
500
%
faster insights
Centralize your quality measurements in one database and get results instantly, not in weeks
1000
%
more feedback
Get categorized, well-structured feedback from linguists that helps you train your engines better
feature
Quick eval with Adequacy-Fluency
Rating Scale approach is the fastest and cheapest way to measure how humans perceive raw MT quality
feature
Detailed eval with Error Annotation (MQM)
Error Typology approach is the best way to deeply analyze mistakes made by an MT engine and get specific ideas for retraining
feature
Post-Editing analysis with Edit Distance
Perform a “mock post-edit” online or import PEMT jobs from your Translation Management System to calculate TER
feature
Unlimited number of user accounts
Our subscription plans do not have any limits on how many linguists or project managers can access the system — bring them all in at once!
feature
Central database of MT quality scores
Collect and manage all results of your human quality evaluations for your Machine Translation engines in one place
integrations

Import machine translations from any TMS

Plays nicely with your existing toolset

We recognize that linguistic quality programs don’t happen in a vacuum: there are all sorts of other localization tools you or your clients are already using in order to produce localized content. This is exactly why we made ContentQuo interoperable and standards-based.
Upload any XLIFF, Excel, CSV, or TMX
All popular XLIFF varities from most popular CAT and TMS tools can be read and written by ContentQuo. Other bilingual formats like XLSX can be read, with per-project column configuration. 20+ file formats!
Import existing offline quality scorecards
Have been doing LQA in spreadsheets for years? ContentQuo can import all of those, so that you don’t lose any of those precious quality KPIs and can centralize them from day 1
Read XLIFF revision history or compare versions
We import change history automatically from supported XLIFF flavors like Trados or memoQ. If unavailable, you can always ask ContentQuo to compare 2 file revisions and find changes
Pull data in and out of your TMS automatically
Through our TMS integrations, we can synchronize your user accounts, projects, and translation jobs into ContentQuo - and also make changes in your TMS when certain conditions are met
Write all modified translations back into XLIFF
For easily updating your TM after LQA or mock post-editing sessions, ContentQuo writes all the changes done to translated text back into your bilinguals so that you can import them into TM
Export scorecards into customizable Excel
When you just need that offline copy, export a single quality scorecard from ContentQuo into an Excel, or export a summary of all your quality evaluations, or get a ZIP with all individual scorecards - in just a click

Infinitely flexible to match your quality process

ContentQuo offers unprecedented flexibility to exactly match your existing human evaluation workflow and methodology. Your team doesn’t need to adapt the way they work - instead, the tool adapts to you!
Customizable error typologies
Mix and match any MQM error categories into your custom quality profile. Define any weights and penalties, even for specific combinations of category and severity if needed.
Customizable rating scales
Use 3-point, 4-point, or 5-point scales. Assess Adequacy and Fluency, or Accuracy / Language / Style, or Expected Edit Effort - the choice is yours. Different projects can use different scales.
Customizable number of evaluators
Get averaged ratings from 2 to 5 linguists simultaneously for more objective Adequacy-Fluency evaluations. Or just assign 1 linguist.
Customizable workflow steps
Have a senior linguist double-check the assessment, or disable post-editing to focus on evaluation only. Switch the workflow per project if needed.
Customizable Edit Distance metrics
We support TAUS Edit Density and CharacTER out of the box and can integrate any custom Edit Distance metrics on demand
Customizable quality grading & scoring
Define any labels for quality score bands, like “PASS, “Excellent”, or “🔥🔥🔥”. Apply the default MQM scoring formula or use programmatic grading for top flexibility.

As safe as a safe. Or even safer.

Whether you’re a corporate team or an LSP, you know that working with proprietary or even high-sensitivity content requires you using only secure and trusted SaaS platforms. This is what we obsess about at ContentQuo - keeping your (and your client’s) data safe.
Secure public cloud hosting in the EU
By default. your ContentQuo will be hosted in a secure Tier4 grade datacenter facility in the European Union (typically Germany). Our preferred hosting partners are Hetzner GmbH and Microsoft Azure.
Private cloud hosting in EU, US, ANZ
For higher security and deeper data isolation, we can deploy ContentQuo in your Microsoft Azure account - our team will manage it for you while you cover the cloud subscription costs.
On-premise hosting in your datacenter
For those rare, highly-sensitive cases where you need 100% control over your infrastructure, ContentQuo can be deployed on premise on your  hardware - even without Internet access.
Stringent security audits by top companies
We have passed numerous information security audits from our customers working both in the commercial and in the government sectors, so that you be confident about your data being secure.
Fully GDPR compliant EU company
As an EU company serving many EU public sector organizations, we place a special emphasis on treating your personal data right, including dedicated features such as user anonymization.
Flexible access and visibility controls
Different global roles have different permissions in the system, while different teams can have access to different sets of projects in the system. Even granular permissions on each workflow step can be set.

Easy to learn, implement, and succeed

Even great software can be useless if you don’t understand how exactly could you achieve great results with it. We work hard to keep our platform as simple as possible over the years, while listening to user feedback and taking it onboard to make everything better for you.
5-minute learning curve for your linguists
If your translators has seen a CAT tool, they will be able to start with ContentQuo in no time. Our built-in guidance explains the core of quality evaluation, rebuttal, and arbitration processes to each one.
Always personal client onboarding
Regardless of the subscription plan you choose, count on having a real human expert introduce you to all the details of using ContentQuo in your specific scenarios and situations
Technical support you can count on
Have a difficult process question, or looking to workaround a limitation? Our team members have decades of localization experience on both buyer and vendor side -- we get it!
Continuous new features and bugfixes
We have been building our technology non-stop since 2015, and our Engineering team ships new things and fix problems every 2 weeks -- more cool stuff is always coming for you to use in your work
Built on best practices for linguistic quality
We happen to work with some of the brightest minds focused on Localization Quality Management in enterprise, government, and LSP sectors. Our tech helps you unlock that know-how.
Start small - or go big on day one
Our flexible product suite and licensing approach means that we can offer the best value to meet your process, scale, and integration needs. This is why ContentQuo is preferred by big and small teams alike.
suppliers

Bring your vendors, your freelancers, 

or  your in-house team

ContentQuo Evaluate MT is not tied to any specific Language Service Provider - you choose your own suppliers! Many LSPs from the Global Top-100 already use our platform. Here are some of them:

RESOURCES

Learn best practices for Machine Translation quality management

webinar
5 steps to take Neural Machine Translation quality under control
INFOGRAPHIC
MT quality evaluation methods comparative map

Is it time to automate your MT human evaluation workflow?

Let’s have a 15 minute conversation and we’ll help you figure this out.
talk to an expert
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.