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 · Sariyer

AI Consulting for Sariyer businesses

When the only thing joining two files is a row number, that join breaks quietly one day: a line is inserted in one, nothing moves in the other, and a wrong pairing travels on unnoticed. On a head office floor around Huzur and Baltalimani it shows up as two spreadsheets open side by side, with somebody reading across them by row. In the tool we wrote for Vipkurum, the answer is written inside the record of the question it belongs to.

Get a quote Tell us about your project

AI Consulting — Sariyer

Sariyer leans on the head office floors along the Maslak plaza line. If your knowledge lives in two files at once, the content sits in one place, its correct counterpart in another, and an index number in between. Revise either file and the link breaks, and the break shows only when somebody notices a wrong result. The work closest to this district's working profile is Vipkurum's question cropping tool, built in an education technology context. Where it does not overlap belongs on the page: a contract archive, version tracking and a payment product are outside that tool, and for banking we have nothing to show. Gemini and OpenAI models do the matching, Python and OpenCV do the cropping, and the records sit in MongoDB.

  • The link built on a row number disappears, because the counterpart from the key is written inside the record of the question it belongs to.
  • Revise a question and its counterpart travels with it, so keeping a second file current stops being a step at all.
  • The scope is written down: the tool was built for education material, and a contract archive or version tracking sit outside it.

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 answer sits in the question's record, not a row number

The question cropping and answer key matching tool we wrote under the Vipkurum roof. The real change is where the link lives: the correct answer is no longer parked in a second file behind a row number, because Gemini and OpenAI models write it inside the record of the question it belongs to. Revise a question and its answer travels with it, so keeping a second file current stops being a step. Python and OpenCV handle the cropping, records sit in MongoDB, and the service reaches Laravel through Node.js. The scope is education material; a contract archive, version tracking or a payment product are not work we can show.

Read the case study →

Working with Sariyer 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

  • We match two spreadsheets by row number and it slips. How do you close that?

    By removing the link. The correct answer does not sit in a second file; it is written inside the record it belongs to. Insert a line and there is no alignment left to slip, because there are no longer two lists to align.

  • If a model writes the counterpart, what happens when it writes the wrong one?

    We quote no accuracy rate, because we hold no measured figure. What the arrangement gives you is this: since the answer sits in the question's own record, a wrong one shows up in a single record instead of being hunted between two files.

  • Our content is internal training material and product documentation, not exam questions.

    The tool was built for education material and we say so plainly. What carries across is not the subject but where the correct counterpart is kept. For a contract archive, version tracking or a payment product we have nothing to show.

  • How much of this work actually goes to a model?

    The cropping side is image processing and no model runs there. A model comes in when the counterpart is attached to its record, and when a question nobody has is written from existing examples. In consulting our first output is usually that division itself.

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