Skip to content
⁨NDDev RnD⁩, home page
Community
Language: English
Display
Theme
Motion

Technical feasibility study

A technical feasibility study at NDDev answers one question: can what you have in mind be built on your data, your equipment and within your constraints. Our RnD practice tests the hypothesis with experiments and delivers a report with measurements, risks and a recommendation. Our team has 100+ projects behind it, and every study starts with criteria you can measure.

Who it is for

Executives, CTOs and product owners before they commit a development budget: when it is unclear whether accuracy, speed or data will be enough for the plan.

  • A conclusion backed by measurements
  • Approaches compared on your data
  • Risks and the limits of the technology
  • A recommendation for the next stage

What the study report contains

A study starts with a question that can be tested: what accuracy, speed or load you need, and on what data. For that question we build a test setup with the same constraints as the real work, take a simple baseline method and compare more complex approaches against it, open-source libraries included.

  • the hypothesis and success criteria in writing
  • a description of the data, equipment and test conditions
  • a comparison table of approaches with measurements
  • experiment code you can rerun
  • a list of rejected methods and the reasons
  • risks, limitations and an estimate for the next stage

There are two possible outcomes. The first is a working method with code and measurements, ready to become a development specification. The second is a proven finding of technical impossibility: the methods tested, the limit of the technology, and what would have to change in the task to make it achievable.

From question to conclusion

You receive interim findings as the work goes on, and the outcome in a single report.

  1. We frame the question

    We write down what has to be proven, which numbers will measure it and what constraints the equipment and data impose.

  2. We build the test setup

    We prepare a representative data set and an environment with the same constraints as the real work.

  3. We compare approaches

    We start with a simple baseline and add more complex methods one at a time, comparing them on the same metrics.

  4. We write the report

    We bring together the measurements, rejected options, risks and a recommendation: build it, how to build it, or change the problem statement.

Questions before a study

What determines the cost of a study?

The number of hypotheses, the volume of data, the complexity of the test setup and the number of approaches to compare. The price block shows what a study starts from. The work is paid as a fixed monthly sum, and you decide whether to continue.

What if the study shows the task is impossible?

That is a result too, and you get it before investing in development. The report gives the measurements, the methods tested and the conditions under which the task becomes achievable: different data, a different accuracy target or different equipment.

What data do you need to start?

A description of the task and a sample of the data, or access to the equipment. If the data is thin, our first step is to assess whether it is enough to test the hypothesis, and we tell you straight away.

Can we order the development from you afterwards?

Yes. The report becomes the basis of the specification, and NDDev.Dev or our AI practice takes on the development as a separate scope. You can also hand the report to your own team: it is written for engineers and contains everything needed to repeat the experiments.

How is a study different from a consultation?

A consultation gives you an engineer's opinion based on your description of the task. A study tests the hypothesis with experiments on your data and gives a conclusion backed by measurements.

from ₽280,000

Free warranty
40 days

Prices are starting prices for information only and are not a public offer. The final price is fixed in the specification.

Research on demand

Name the idea you have doubts about

Describe the task and its constraints and attach a sample of the data. We will suggest which hypotheses to test first.

We reply within 24 hours

Assistant

Loading the assistant…