EN TR
Get a quote
View all services
Web software

One app carries the brochure pages and the request queue.

System architecture

The pieces and their boundaries are settled before code.

Enterprise & commercial software

Several institutions, one system, records kept apart.

Mobile app

One codebase, both stores: iOS and Android.

Software consulting

We take a running system over and improve it.

AI consulting

AI only where the problem actually calls for it.

Get a quote
EN TR

AI Consulting · Ankara

AI Consulting for Ankara businesses

When two documents point at each other, whoever holds them side by side has become part of the system. If the division of labour on the institutional floors around Oran looks like that, a training unit prepares the question, an assessment unit keeps the key in a separate file, and the pairing rests on somebody's attention. Put that way, AI consulting in Ankara is not a question about picking a model. It is a question about which record the answer lives in. In the tool we wrote for Vipkurum, the key stops being a separate document.

Get a quote Tell us about your project

AI Consulting — Ankara

Ankara reads as a management layer rather than a production one: head offices, institutions split into many units, in-house training and certification programmes. If your question bank does not sit in one pair of hands, the unit that writes and the unit that marks are working from different folders. The closest work we have is Vipkurum's cropping and matching tool. Where it stops is clear: it was written for exam content, and we have nothing to show for internal policy documents, regulation archives or contract libraries. There is no public-sector tender in the portfolio either. Python and OpenCV crop, Gemini and OpenAI models match, records sit in MongoDB, and Node.js joins the service to content management on Laravel.

  • The answer from the key is written into the cropped question's own record, so the intermediate step of holding two files together disappears.
  • A question you do not have is written from the examples you do have, and its key is produced in the same step rather than looked up later.
  • What sits outside: internal document search, regulation summarising and document classification are not part of this scope.

Why Algoriel

  • The image does the reading

    Vipkurum reads optical form marks with OpenCV from the image of the sheet, not through the optics of a reader appliance.

  • Cropping apart from judgement

    Image processing decides where a question begins and ends; the language model only binds the answer to it. The two jobs never mix.

  • Model output lands in a record

    Vipkurum writes the answer key into the cropped question's own record instead of keeping it as a separate document.

  • A person has the last word

    At Tahlil.com the model drafts the reading and a doctor approves what reaches the patient; an unapproved reading never appears on screen.

What we build

AI Consulting

  • Where no model is needed

    Marks on an answer sheet are read from the form image itself; the threshold that decides whether a bubble counts as filled sits in the code, and no language model runs in that step.

  • Splitting the job in two

    Computer vision sets the question boundaries; a language model then ties each answer in the key to its question. Cropping accuracy never rests on a model's interpretation.

  • Removing the hardware step

    Marking stops depending on a reader device; when the form layout changes, we adjust the reading service in software instead of ordering new hardware or device-specific forms.

  • Turning existing material into a pool

    A booklet page is split into individual questions, the answer key is written onto each question's own record, and the questions already on hand become the pattern for new ones.

  • Tying the output to a person

    At Tahlil.com AI prepares the first reading, a specialist physician writes and approves the interpretation, and responsibility for the text that reaches the patient stays with the physician who signed it.

  • Queueing the generation

    In Secretum text and images come out of two separate generation paths; image generation is not instant, so that work goes onto a queue, and the finished file is stored once instead of being regenerated in the feed. Secretum is the product we are building now.

Work like this

AI Consulting · Case study

Vipkurum question cropping — the key leaves the file and settles in the record

The cropping and answer key matching tool we wrote under the Vipkurum platform. Once a booklet page has been separated into single questions, Gemini and OpenAI models bind the answer from the key to the question it belongs to; the answer is no longer a line in another file but a field of the question's own record. Holding two documents side by side, and checking the pairing by eye, stops being a step. A question that is missing gets written from existing examples, with its key produced alongside. Cropped images and question records sit in MongoDB, and the service reaches Laravel content management over Node.js.

Read the case study →

Working with Ankara teams, remotely

We have no office, so work for this city runs remotely and we do not offer on-site support. What was done, and why it was done that way, stays in writing. The project pages mark which builds came out of a partnership.

  • The first release ships as a working piece; extras follow later.
  • Every client we know of is in Turkey.
  • Decisions stay in writing; progress shows in working screens.

Frequently asked

  • One unit writes the questions, another prepares the key. Does that split survive in software?

    The order of work does not change; what changes is where the answer is stored. Instead of waiting in a separate file, the key is written into the relevant question's record, so both units look at the same place.

  • Do you have a public-sector reference?

    No. There is no public tender in our portfolio and we do not imply otherwise. What we can show is the tool written under Vipkurum's education platform, and that is what you would judge us on.

  • Do we have to trust the questions a model produces?

    No. The model writes a new question and its key together from the examples already in hand; the institution decides which of them enters the bank. Generation is a suggestion layer, not an approval.

  • Our booklets are printed and we hold no digital copy. What then?

    An image of the page is enough. The tool works over the scanned form of printed material, so the content does not have to be typed up as digital text first.

AI Consulting in other cities

The other cities where we publish this service are below, grouped by province.

Other services for this location

Let us look at which step needs a model and which one needs computer vision.

Get a quote or open the service