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 · Beylikduzu

AI Consulting for Beylikduzu businesses

What gets shown sits on a screen; what actually exists is bound. A spiral file is handed over whole, while the person in front of the display is after one item inside it. Between the two lies the scanner glass, and that is where the real decision is made: does the page stay a single image, or is it separated into records? In the tool we wrote for Vipkurum we took the second route, and a printed booklet became a pool where every question can be called on its own.

Get a quote Tell us about your project

AI Consulting — Beylikduzu

Beylikduzu leans on showrooms and sales floors, where a product meets its buyer on a display. If the material you have accumulated stays printed and bound, a gap in resolution opens between the two. The portfolio work closest to that fabric is Vipkurum's question cropping tool, and its reach is stated plainly: it was written for education technology. Product catalogues, sales sheets, furniture and textile work are not things we can show, because no work from those sectors exists in the portfolio. What carries across is the resolution of a record: not the whole document, but a single item inside it, callable on its own. Python and OpenCV handle the cutting, Google Gemini and OpenAI models the matching, records sit in MongoDB, and the service reaches Laravel content management through Node.js.

  • A printed page does not remain one undivided image; every item inside it becomes a record that can be called on its own.
  • Because a single item is callable, the same material can be arranged again without preparing the source from scratch.
  • The limit is stated: this separation was built on education content, and we have nothing verified on product catalogues or sales sheets.

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 — a bound booklet became individually callable records

Vipkurum works in education technology. The material at hand is a printed question booklet: Python and OpenCV separate a page into single questions, and each question becomes its own record. Those records and the cropped images sit in MongoDB, while the Python service reaches Laravel through Node.js. The booklet stops being a file that can only be handed over whole: one question inside it can be called, and the same material can be arranged again. Gemini and OpenAI models take on the matching and write new questions from existing ones. The scope is education content; a catalogue or a sales sheet is not this tool's territory.

Read the case study →

Working with Beylikduzu 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

  • Could you make our bound catalogues searchable?

    We have no work to show for that, so we do not promise it. Our evidence sits in education content: a printed booklet was separated into single questions and each question moved into its own record. Promising the same for catalogue material would mean describing work that does not exist.

  • Instead of separating items, why not just show the whole page?

    Because then the page becomes the unit you call. Needing one item means carrying the whole page, and arranging the same material in a different order stops being possible. Drop the record's resolution from page to item and both constraints go at once.

  • We have no scanner and photograph pages on a phone.

    At Vipkurum we worked from scanned pages. A phone photograph is not the same condition; skew and shadow land directly on the cutting step. Before promising anything we measure it against a sample you send, and we report what we find as it is.

  • Would you not need to walk our sales floor?

    We keep no office, work fully remotely, and do not offer on-site support. What is under discussion is not the floor anyway, but the unit your printed material should be separated into. We also say up front that we have no client on this district's furniture or textile side.

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