New: Claude services for organisations See the details →
Lab 05 · R&D
What GET builds and publishes: open source, beta products and research.
Our public repository, products in market testing, prototypes and research on Arabic and regional AI.
The lab at a glance
Each lab decides on the business side and delivers on the technical side. Select a service to read what it covers.
Services
Scope, responsibilities and acceptance criteria are agreed before any engagement begins. Fixed prices are shown for defined scopes; everything else is Per quote, with the right team for your project and budget.
Arabic and bilingual platform evaluations, a MENA AI regulation tracker, dated briefings and adoption research.
Public repositories of templates, skill libraries, test kits and evaluation harnesses at github.com/GET-AI-MENA.
Products in market testing, each with its status, early-access sign-up and a feedback route.
Fast feasibility prototypes and 12-week MVPs that end in a go or no-go.
Prices are in US dollars and exclude VAT. Fixed prices cover a defined scope, confirmed in writing before work starts. Payment: under $2,000 in full at the start; from $2,000, 30% to start, 40% at mid-point and 30% on delivery; monthly services are billed monthly in advance; 14-day terms. 50% of the GB2 fee is credited against a follow-on engagement started within 90 days (up to 25% of that fee), and the TT3 discovery sprint is credited against its build. Licences, cloud, inference, channel and exam fees are never inside a fee.
Send an enquiry
Fill in this short form. Our team reads every enquiry and replies by email.
Not sure which service fits? Find your starting point · See all six labs
Our team will review it and reply to the email address you provided.
Insights
Which AI rules bind in Saudi Arabia, the UAE, Jordan, Qatar and Egypt as of October 2026, which only guide, and a monthly routine to track what changes next.
Read the article ↗
A one-page AI proof of concept charter to copy: hypothesis, success and kill criteria locked before work starts, data allowed, time box, decision owner and go/no-go date.
Read the article ↗
A practical method to test whether a model or AI assistant works in Arabic: your own tasks, dialects and Arabizi, normalised scoring, a calibrated judge and a release gate.
Read the article ↗