Me.
My name is Randall Helms, the founder of FFP One, a future finance platform (hence the name). Welcome! I'm very glad you are here and I'd love to introduce you to what I've been working on.
Normally people write blogs like this in the third person, but I'm a solo founder. This is my product, so I'm not going to pretend it's someone else writing this. This is me and these are my words and thoughts.
Although I'm a first-time founder, I'm a fifteen-year veteran of the data world. At the start of 2026, I decided that I wanted to put that expertise to good use by building a product of my own.
That product is FFP One.
What is FFP One?
FFP One is a financial research platform covering equities and equity-focused ETFs and mutual funds, built for anyone who is interested in doing serious financial research on their own, whether they are professional or retail investors.
I'm a realistic man - I'm well aware there are tons of companies out there with a similar pitch, so what makes FFP One special?
The specific thing that makes FFP One tick and (hopefully!) makes it stand out against the competition is the FFP One data layer. I've taken my many years of experience leading teams building professional-grade data systems and used that to build it: it takes massive, unstructured public financial documents, such as 10-K annual statements, and rebuilds them into a clean and well-structured database. The FFP One data layer is the core of the product: it's what powers interaction with AI models as well as more mechanical research.
My working hypothesis is that professionally useful AI systems require the highest-quality data - you will get better answers if the underlying data is impeccable. FFP One is about picking the right tool for the right job; if the data is available right there, already structured and ready to use, the AI model can give you a world-class answer.
The result is a product that I believe is a great offering for anyone interested in financial research: equities research that is fast, affordable, and above all, trustworthy.
Being trustworthy is my obsession
Trust is a hugely important part of what I'm offering you. I've built a system that is able to trace the provenance of any claim it makes all the way back to the original source, i.e. the specific section of the specific filing.
From talking to professional investors about how they use AI, one of the key challenges I've discovered is when to trust the model's output. It always sounds confident, right? These large language models always sound so sure of themselves, but it can be very difficult to judge the accuracy of a number or a purported fact. Where does it come from? Is it directly from an official company source? Is it from a news article summary somewhere? Is it from the model's training data? Or is it a hallucination pulled from the digital ether?
FFP One is different. Whenever you ask a chat question or build a deep-dive research paper, the AI models are working directly against the high-quality data layer that I've built. It's not trawling the raw filing text fresh, doing random web searches, reaching into training data, or, well, just making stuff up out of thin air ... it's pulling answers from a battle-tested database that includes summaries as well as the original filing text. And where FFP One does reach beyond the filings, it's clearly signposted, so you always know what came from the primary source.
If the biggest weakness of doing financial research with a commercial AI system like ChatGPT or Claude is getting 'confidently wrong' answers that sound great but fall apart with a bit of due diligence, FFP One is built specifically to avoid that. If the system tells you that for Nvidia in FY 2026 'Data Center computing revenue grew 59% driven by Blackwell platform demand', then it can trace that claim back to the exact point in the exact document it came from (in this case, Item 7 - MD&A from the FY 2026 10-K filing).
Every answer is traceable back to its source.
What can you do with FFP One?
FFP One allows you to explore company narratives, financial drivers, market performance, and, for funds, look through to understand their components - what industries, what risk factors, what characteristics. My technology turns questions like 'which funds have particular exposure to AI risks?' into a simple database search.
FFP One allows you to look through a fund into its component parts.
FFP One gives you several ways to put AI to work on your research: chat, memos, and, most excitingly, the ability to generate deep-dive research papers.
I will share more information about the feature set soon, but I want to say a quick word about Research Mode. This feature makes it possible to run the most advanced AI models against the FFP One data layer to build deeply researched, scrupulously accurate research papers; imagine having your own on-demand junior analyst ready to go at any time.
Research Mode is FFP One's flagship feature - it's not afraid to push back on you.
What does the product cover right now?
At launch the product is going to cover only companies registered with the US Securities and Exchange Commission (SEC), with the primary data sources being different SEC filings like the 10-K (annual), 10-Q (quarterly), 8-K (one-off releases), and 20-F (foreign filers annual). Additionally, FFP One aggregates and processes data from other sources to cover fund holdings, macroeconomic conditions, foreign exchange (FX) rates, and equity and index price movements.
What I'm really excited about is that this architecture can cover equities from all across global markets - I'm starting with the US but by no means stopping there!
Imagine a scenario where you are a fund manager in Miami and you want to research Italian stocks - today you would have to rely on third-party sources, but soon you will be able to ask FFP One directly in English to walk you through exactly what a given firm says about itself.
Last thoughts
What makes FFP One so exciting is not just what it can do today, but what it makes possible in the future. This is a product that is going to rapidly improve in the coming days, weeks, and months.
This has been a crazy year for me personally. I had this idea on January 12th as I was walking down a staircase in Savannah, Georgia, and here it is as a real, working product (albeit not yet publicly available).
FFP One is a totally bootstrapped product: no investors, no funding rounds, just me dipping into my savings and staring at Claude Code for hours every day and making stuff happen. I'm very proud of the work I've done; doing something of this complexity completely on my own has only been possible due to the revolution in agentic coding that has taken place over the last 12 months. It's allowed me to build something that is a real, full-fledged product, far beyond my own particular expertise in the data realm.
I'm very excited about this product and I'm very much looking forward to seeing real customers using it! I want it to be something genuinely valuable to investors - something that makes the daily equities research workflow easier, more pleasant, and even (hopefully!) fun. If you're interested, please sign up for the mailing list, and I'll let you know as soon as it's ready for customer usage.