FlexSoft Promises Zero Time-to-Market in Fintech Through AI-Native Platforms

Arkady Lobas, CEO of the IT company FlexSoft, spoke at the anniversary Fintech X10 forum. He said that AI-Native architecture and a platform approach deliver zero Time-to-Market in the financial industry. In other words, launching a new product should take practically no time. It makes for a striking line. But the available materials contain no "before" and "after" timelines, no bank names, and no description of the technology. So for now this is a vendor's thesis, not a measured result.

What exactly is being claimed

Two elements are visible in what has been published. The first is AI-Native architecture. This usually refers to systems in which models and agents are built into the development process and the product's operation itself. AI is not bolted on top as a chatbot; it takes part in generating configurations, code, tests, and documentation. The second element is a platform approach, in which products are assembled from ready-made, reusable components rather than written from scratch for each task.

The pairing is logical. The platform supplies a set of building blocks, and AI speeds up assembling and configuring them. But this is a generic scheme, not a disclosed design of any specific FlexSoft solution.

• Claimed: zero Time-to-Market in the financial industry.
• Venue: the anniversary Fintech X10 forum, 2026.
• Basis: AI-Native architecture and a platform approach.
• Not disclosed: the models used, cloud or on-premises deployment, implementation timelines, clients, economic impact.

How to read the word "zero"

A literal zero is unattainable in finance. A new credit or payment product goes through legal review, risk assessment, integration with accounting systems, testing, and regulatory requirements. No architecture eliminates these stages.

The realistic reading is this: the technical part of a launch shrinks to configuration. A product manager describes the terms, the platform assembles the process, and AI prepares tests and documentation. The bottleneck then becomes approval, not development. That is valuable, because at banks development and integration today take up a noticeable share of the launch cycle. But here "zero" is a marketing label for maximum possible acceleration, not a physical quantity.

Where the claim runs into reality

The AI-Native approach has known limitations in finance.

• Explainability. Regulators and auditors require an understanding of why a system made a decision. For generative models this is harder than for deterministic rules.
• Generation errors. Code or configuration produced by a model has to be verified. If verification is manual, the time savings shrink.
• Data perimeter. Financial institutions in Russia operate under strict requirements for data storage and processing. This often means local models and compute constraints.
• Platform lock-in. The more deeply products are tied to a single platform, the more expensive it becomes to switch vendors.

It is the answers to these questions that separate a working architecture from a presentation. Judging by the available text, they were not made public at the forum.

What questions to ask the vendor

If you are a bank, an insurer, or a fintech company considering such offers, do not settle for the slogan. Ask to see specifics.

First: a measurable case with numbers. How many weeks a product launch took before the platform and how many after, and for what type of product.

Second: where in the chain AI sits and who is responsible for the results of its work. You need to know whether a human is in the verification loop and how the model's decisions are recorded.

Third: which models this runs on and where they are deployed. Cost, security, and compliance depend on it.

Fourth: what happens if you leave the platform. You need export formats, rights to the configurations, and migration timelines.

Until FlexSoft shows numbers, it is reasonable to treat the promise of zero Time-to-Market as a direction of travel, not a guarantee. A pilot on a single product with measured timelines will give you more information than any forum speech.