For decisions where there isn’t one yet.
Owners, boards and investors bring me decisions that don’t fit their existing tools. I build what’s missing: the model, the protocol, the market data, the criteria. Then they decide.
When there was no map
In 2018 nobody in Russia had written a clinical trial protocol for CAR-T cell therapy. In 2020, there was none for HER2 bispecific antibodies.
So I wrote the first ones. From scratch.
Notes. BIOCAD, Russia’s leading biotech company. The oncology portfolio: 14 trials, phases I–III, for the EAEU, EMA, FDA and other regulators.
A marine robotics company had to choose: build a 3,000 m underwater vehicle or a 400 m one. Public estimates of how many such vehicles work in Russia differed tenfold: from 50 to 700.
So I stopped choosing between estimates and built the registry: every operator and its vehicles, public tenders since 2020, customs declarations.
Market: ~₽7B in 2025, ~₽10B by 2030. The model went to the CEO and into the company’s reply to the Ministry of Industry and Trade.
Notes. Sources: EIS procurement 2020–2025, Federal Customs declarations, SPARK. The top-down estimate is cross-checked bottom-up.
A European biotech company was raising a Series A to scale up manufacturing. The science was sound. The investors couldn’t price its risk.
So I translated the risk into their language: the financial model, the due diligence, the investor conversations.
Round closed: €1.5M.
When an investor in the Leningrad Region asked a routine question, government bodies and utilities had up to 30 days to respond.
So we built the Investor’s Personal Account: the digital front door of the region’s investment agency, one route for every request.
Response time: 30 → 7 working days.
Notes. Public figure, reported by the regional government and Kommersant. The agency accompanies about 300 projects worth ~₽7 trillion.
The St Petersburg government needed to know how digitally mature the city’s key enterprises were. No instrument existed.
So I built one: metrics, reporting, a transformation roadmap.
The directorate’s annual KPIs: exceeded by 20%.
A biobank of clinical trial samples: valuable in theory, useful to no one.
So I initiated a federal programme to sequence it: ₽30M of state funding.
It became a patent: a diagnostic tool that predicts which melanoma patients will respond to anti-PD-1 immunotherapy.
Notes. Patent RU 2771080 C2, 2022. Public record.
Also in the specification
| Pos. | There was | So I built | Result | Year |
|---|---|---|---|---|
| 1 | No export route: sanctions had closed the usual one. | Another route for an industrial automation manufacturer: 13 product groups, 12 countries, prices benchmarked against Schneider, ABB and Siemens on 150 items, entry through the state export agency. | Buyer search launched in every target country | 2024 |
| 2 | No way to buy the drugs: a clinical standard blocked them. | An expert council of key opinion leaders; lobbied the Ministry of Health to change the standard. | Drugs entered the federal procurement programme | 2020–21 |
| 3 | No way to compare products. | Evaluation standards (PLM, KPI) and DCF models for the board. | +35–50% upside from new product lines estimate | 2025 |
| 4 | No single logic to compare investment initiatives. | One evaluation framework for a family-office-style holding. | The shareholder decides on verified data | 2026, now |
| 5 | Too many candidate molecules, one development budget. | Expert screening on the early development committee. | More than half of the drugs in development passed my screening | 2018–21 |
Method
Before anyone decides, one question:
Most of the variables won’t matter. Three or four will. Finding them is the work.
About
The industries changed. The situation didn’t: an important decision, and no map for it yet. That’s where I go.
I don’t make the decision. I make it possible.
Contact
ax2207@yandex.rulinkedin.com/in/ialvovsky
If it’s already obvious what to do, you probably don’t need me.