Founders' perspective
Why we built AlphaPulse
A founder's account of information overload, market psychology, and the decade-long journey to turn raw world events into investment signals.
The problem we couldn't stop thinking about
Financial markets have always been a game of information. But somewhere in the last decade, the rules changed. The volume, velocity, and noise level of information reaching investors crossed a threshold where human processing simply can't keep up.
Pump-and-dump schemes. Hypes. Crowded trades. FOMO. These aren't just market dynamics - they're symptoms of investors being overwhelmed, reacting instead of deciding.
The question that kept us up at night wasn't “how do we predict markets?” It was simpler and harder: how do you efficiently compress the world's information, sort out what genuinely matters, assess its real impact, and act - or knowingly hold back - before the crowd does?
That became our Holy Grail. A tool that doesn't just aggregate news, but translates the global information flow into structured, comparable, time-stamped signals. Something that lets you see risk/reward clearly before everyone else has priced it in.
We started collecting data before we had a company
Long before Alpha Data Analytics was formally incorporated in early 2022, we were already building pipelines. Painstakingly collecting, curating, processing, and storing world events data - because we knew that any serious signal intelligence system lives or dies by the depth and quality of its historical record.
You can't train models on data you don't have. You can't backtest signals against events you didn't capture. So we started early, deliberately, knowing the dataset itself would become one of our core assets.
Three perspectives, one obsession
The founding team came together around a shared frustration with how fragmented and reactive market intelligence tools were - but each of us brought a different lens to the problem.
Anders Nygaard - Institutional Know-How
Anders brought decades of institutional investment experience - the practitioner's understanding of how real capital decisions get made, what information desks actually need versus what they get, and where the gaps between research and execution create risk.
Natallia Gorbunova - Engineering Excellence
Natallia shaped the architecture that makes it all possible - reliable ingestion at scale, clean data pipelines, and a platform that can process millions of events without losing signal fidelity. Engineering excellence isn't a nice-to-have in this domain; it's the product.
Vadim Skritskii - AI Research & Expertise
Vadim's decade of AI research informed the core signal models - how to extract meaningful structure from unstructured text at scale, how to represent sentiment with statistical rigour, and how to make the system adaptive without making it fragile.
Hundreds of conversations, one pattern
Before we wrote a single line of product code, we talked to people. Hundreds of them. Individual traders, institutional investors, conservative regulated funds, portfolio managers, hedge funds, quants - but also journalists, airline revenue managers, government officials, AI researchers, data engineers, consultants, and marketing specialists.
The diversity was deliberate. We wanted to understand not just how investors use information, but how modern information flow shapes perception across every domain where decisions get made under uncertainty.
One pattern emerged everywhere: the problem wasn't access to information. It was the inability to separate signal from noise at the speed that markets demand.
Those conversations didn't just validate our thesis - they sharpened it. They showed us exactly which dimensions of market intelligence mattered most, and which “features” were really just noise dressed up as insight.
The pivotal moment: LLMs changed everything - twice
The proliferation of generative AI and large language models (LLMs) was a double-edged shift.
On one side: the information problem got dramatically worse. LLMs enabled the production of financial commentary, analysis, and narrative at a scale and velocity that no human team could ever read, let alone evaluate. The noise floor rose sharply.
On the other side: the same class of technologies made it possible, for the first time, to process that information at comparable scale. To read everything, weight it, and surface only what genuinely moves markets - automatically, continuously, in near real-time.
We had spent years building the data foundation. LLMs gave us the engine. AlphaPulse became possible at exactly the moment it became necessary.
From the signal to the screen
On 14 January 2025 Jakub Hankiewicz joined as Head of Digital Assets product, responsible for the b2c products. He had been with us from the early beginnings as a ux/ui designer, shaping how a signal reads on screen - which is how he came to the role. He now also actively supports shaping the b2b products and is involved in Adalytica's strategy.
