Agentic AI services
AI agents and AI features, proven before they ship.
Fikrano designs, builds and tests AI for companies. We also test the AI that teams build themselves, so every release goes out with evidence.
Already shipping AI? Get a free AI Quick Check
Our services
- AI agentsAgents that do multi-step work with your tools and data, within limits you set.
- AI featuresAssistants, search and drafting, built into your product.
- AI evaluation & QATest suites and release checks for AI you build yourself.
- Software QARelease testing and automation for the rest of your product.
- Arabic & bilingual AIBuilt and tested for dialects and Arabic-English mixing.
How we work
One method for everything we build or test. Testing runs through every step, not only at the end.
01Design
We map the job, the risks and what a good result looks like. Then we agree the scope and a fixed price.
02Build
We build the agent or feature in your stack, with the tests written alongside it.
03Prove
Every change is scored against a test set before release: quality, safety and the cases that matter to you.
04Run
After launch we keep testing each change, and report what changed and why.
What you get
Three of the documents we hand over. Each one is written so that your team can read it in minutes.
- Example
Evaluation report
A score for every release, with the failures, the evidence and the fixes.
See how we work - Example
Agent design sheet
What the agent may do, where a person approves, and what gets logged.
See how we work - Example
Release check
A gate in your pipeline that stops a release when a check fails.
See how we work
Why Fikrano
Specialists in AI that has to work.
Fikrano is led by an AI QA team lead who spends every working day evaluating language-model features, retrieval systems and agents. The tests we run on our own builds are the tests we run on yours.
Evidence, not opinions
Every release gets a score against a test set you can read.
Fixed scope, fixed price
You know the cost before we start. No hourly surprises.
Your stack, no lock-in
We build and test with your tools or open source, and you own the deliverables.
Hours that overlap yours
Based in Cairo, with working hours that overlap Europe, the Gulf and the US East Coast.
Questions we get asked
Short answers. For anything else, email us.
Do you need our code or production data?
No. We work in a test environment, with accounts you provide. An NDA and a data-processing agreement are standard.
Which tools do you use?
Yours, or open source. For evaluation that means your existing platform or frameworks such as promptfoo, DeepEval and Ragas. For test automation, Playwright and your CI. We never lock you into tools of ours.
Can you build the agent and also test it?
Yes, and that is the normal case: the tests are written alongside the build and run on every change. If you would rather build in-house, we test what you build.
What happens when we switch models?
That is when a test set pays off. We run your suite against the new model and show you what changed before you switch.
Do you guarantee bug-free releases?
No one honestly can. We give you evidence and a go or no-go recommendation. The release decision stays yours.
Where are you based?
Cairo, Egypt. We work with teams in Europe, the Gulf and North America, in their working hours.
How do we start?
Email hello@fikrano.com with a few lines about what you are building or shipping. If you already have AI in production, ask for a free AI Quick Check.
Founding clients
Three founding places.
We are taking three founding clients in our first months. Each gets a lower price on their first project, in exchange for a case study written together once the work is done, and published only with your approval.