Trisigma to present a case study on rolling out a Russian A/B testing platform at CNews St. Petersburg

On 15 October 2026, Maria Sapelnikova, a business partner at Trisigma, will speak at the conference "CNews St. Petersburg: Advanced Technologies and Cases of Russia." She will describe the experience of deploying a Russian platform for A/B testing. The announcement CNews published on 13 October is short: it names no customers, gives no project timelines and cites no measurable results. Those details will have to come from the talk itself.

What is known so far

The announcement says little, but what it does say can be verified:

• Date: 15 October 2026.
• Venue: the conference "CNews St. Petersburg: Advanced Technologies and Cases of Russia."
• Speaker: Maria Sapelnikova, business partner at Trisigma.
• Topic: deploying a Russian platform for A/B testing.
• Format: a practical case study, not a product presentation with a roadmap. That is how the talk's title frames it.

The announcement does not mention the client base, the number of concurrent experiments, the load or the architecture of the solution. Retelling what the source does not report would be speculation.

Why the market needs a domestic experimentation platform

A/B testing has long been a working tool for product teams in e-commerce, banking, telecom and media. Any change to an interface, pricing or recommendation logic is tested on a portion of the audience and compared against a control group.

Since 2023, the Russian market has had another reason to look at local solutions. Google shut down Optimize on 30 September 2023, and foreign SaaS experimentation platforms remain in question because of sanctions risks, payment issues and the storage of user data. That has created demand for platforms that can be deployed in a company's own environment and integrated with internal data warehouses.

AI adds a separate angle. Teams that ship recommendation models, ranking systems and LLM assistants increasingly validate them through experiments, because offline metrics are poor predictors of how real users behave. An A/B testing platform becomes part of the MLOps infrastructure. The announcement does not make clear how far Trisigma is tied to this scenario.

What to look for in the case study

If the talk is substantive, it could shed light on several factors that usually determine whether a deployment succeeds.

First, integration. How long did it take to connect to existing event sources, the data warehouse and the analytics system, and did the team have to rewrite its logging?

Second, statistics. Which methods does the platform use: the classic t-test, sequential testing, CUPED and similar variance-reduction techniques, multiple-comparison control? These choices determine how quickly a team gets a reliable result and how many false wins slip into a release.

Third, organization. Who owns the platform: product analytics, a platform team or the business units? How is the process set up for approving hypotheses and reviewing failed experiments?

Fourth, the cost of switching. What happened to historical data, and how were results compared between the old and new tools?

How to use the announcement in practice

Anyone choosing or replacing an experimentation platform should come to the talk with a ready list of questions:

• what traffic volume and how many parallel tests the platform handled in production;
• where the data is stored and whether the solution can be deployed in an isolated environment;
• how metrics are implemented: a ready-made catalog or custom calculations on the customer's side;
• how long the deployment took from pilot to regular use;
• what effect the customer got, and whether it was measured by the experiments themselves.

If the speaker answers the last two questions with numbers, the case will be useful. If not, it remains a story about installing a tool, not about what that tool delivered for the business.