Eastern Pine Ventures / Accountable intelligence systems
Markets move quickly, and good reasoning is easy to lose. Eastern Pine Ventures builds tools that keep the evidence, the decision, and the outcome together so people can return to the work and learn from what happened.
01 / Mission
A market idea can now pass through a filing, a post, a model, a chat room, and a trading venue in minutes. By the time it reaches somebody, the useful context may already be gone. Our tools are meant to keep that context close.
Models can help with the work, but people remain responsible for the decisions.
That trail matters because a result by itself tells us very little. We need to know what somebody saw, what they believed, what could have proved them wrong, and what really happened.
02 / Market history
The old photographs make price discovery feel physical: boards, paper, telephones, messengers, crowded floors, and people trying to keep up. The feed was slower, but the basic problem was familiar. Figure out what matters before somebody else does.
The market is not going to slow down, so the record has to keep pace with the machinery.
Why this history belongs here
These photographs make market work visible: the machinery, the delays, the judgment, and the people trying to keep the record straight.
A short market history
Read the price move here. Open the museum only when you want the full story.
South Sea shares surged during 1720 and then collapsed. Fortunes built on the run-up disappeared, and the crash forced a rescue and restructuring of the company.
The Dow rose about sixfold from August 1921 to September 1929. It fell nearly 13% on October 28 and nearly 12% the next day. By July 1932 it was 89% below the peak and would not regain that high until 1954.
The official gold price had been $20.67 an ounce. Beginning February 1, 1934, the Treasury bought gold at $35 an ounce. That was a rise of roughly 69% in gold's dollar price and a deliberate devaluation of the dollar against gold.
Technology reduced some trading costs and helped quotes update faster, but it also gave speed a price of its own. The dot-com cycle showed the other side: the Nasdaq-100 peaked near 4,705 in March 2000 and bottomed near 805 in October 2002, a decline of more than 80%.
U.S. home prices fell about 30% from mid-2006 to mid-2009. The S&P 500 fell 57% from its October 2007 peak to its March 2009 low.
The Dow's intraday decline reached 9.16%. It dropped another 573 points in five minutes, then recovered 543 points in roughly a minute and a half. Some stocks briefly traded for a penny.
03 / Systems
FoxClaw, Planifier, and CoinFox each handle a different part of the work. What connects them is the record: where an idea came from, how it changed, who approved it, and how it turned out.
Public accountability
CoinFox is where the work becomes public. People will be able to publish market ideas, show their reasoning, question one another, and come back later to see what held up.
Models can propose. People decide. The reasoning and the result stay attached.
02 / Private infrastructure
FoxClaw is the private intelligence system behind the work. It connects research, models, tools, rules, and human direction without quietly handing the final decision to a model.
The working memory.
03 / Planning layer
Planifier takes a trade or market idea and gives it enough structure to question: the thesis, evidence, risk, timing, confirmation, invalidation, and the review afterward.
The plan you can inspect.
04 / Public layer
CoinFox gives market ideas a public memory. The claim, evidence, discussion, corrections, and outcome stay connected instead of disappearing into another feed.
The public record.
How they connect
FoxClaw
private research and working memory->Planifier
turns the idea into a reviewable plan->CoinFox
opens the claim and outcome to public review04 / Roadmap
Build trustworthy intelligence infrastructure that earns greater capability through evidence, testing, and real-world use.
The roadmap is based on proof, not hopeful dates. A stage moves forward when the work meets its gate, not when a quarter ends.
We label the work plainly: proven, in build, planned, or research.
In build
Private market-intelligence infrastructure that keeps context, limits, and decisions connected.
Private reconstruction
The public place for market ideas, discussion, corrections, and outcomes.
Active preview
Educational planning, journaling, paper practice, and review.
How we build
The current job is to get one supervised paper workflow running reliably from start to finish, with enough context to explain and repeat it.
Exit condition: A complete supervised paper workflow can run repeatedly, explain what happened, and preserve the evidence needed to reproduce it.
Make the underlying system understandable to somebody new: clear setup, safe startup, useful demonstrations, export, and a Control Center that explains its own state.
Exit condition: A new user can install FoxClaw, understand its state, complete a supervised paper workflow, and inspect why the decision was made in under ten minutes.
Put FoxClaw, Planifier, and CoinFox in front of people outside the project and learn whether they return, understand the output, and find it useful.
Exit condition: Real users repeatedly obtain useful, reproducible intelligence from the products.
Broaden coverage, APIs, contributor tools, and infrastructure only when the earlier work gives us a reason to.
Exit condition: Expansion is justified by operating evidence, not by ambition alone.
This is the public map: direction and proof. Private architecture, live execution plans, sensitive controls, internal source history, and unresolved research stay private.
05 / The record
This is a small running log of decisions, lessons, and changes in direction. It is here to show the work, not to manufacture momentum.
Eastern Pine Ventures was reframed around one connected market-reading system: FoxClaw, CoinFox, and Planifier.
Company
A portfolio-level contract model was recorded so products can send reviewed state and outcome receipts back to FoxClaw without unsafe control.
Architecture
The rise of agentic AI sharpened the question: how can a system help with complex market decisions without pretending the model is in charge?
Origin
06 / Founder's note
I was born in Concord, Massachusetts, in 1994. My family moved to New Hampshire when I was seven or eight, and I have lived here ever since.
Most of my working experience came from the printing field. At Spectrum Marketing in Manchester, I ran a sheet-fed offset press and learned how much has to go right before a finished piece ever reaches somebody's hands. A press is a whole system of moving parts, material, timing, judgment, and inspection. If you miss a small problem early, it can travel through the entire job.
I did not realize it then, but that way of thinking stayed with me.
When I was nineteen, I was in a serious car accident and had my hip reconstructed. Years later the reconstruction failed. I left printing in August 2023 because I could barely walk, and for a long stretch I was considered disabled. Finding the right surgeon while dealing with insurance and money took much longer than it should have.
Eventually I was able to have a total hip replacement at Dartmouth Hitchcock Medical Center in Lebanon. They took great care of me. I feel better than I have in years and barely limp now. These days I work in tree removal for a company in Massachusetts while I build Eastern Pine Ventures.
I have wanted to start a company for as long as I can remember. Near the end of 2025, agentic AI made the idea feel possible in a new way. FoxClaw started with a pretty simple thought: could I build my own finance-focused version of an agent, one that remembered the owner's direction and could help make sense of complicated market decisions?
I found out quickly that the hard part was not getting an AI to produce an answer. The hard part was keeping the evidence, limits, authority, and outcome attached to that answer. FoxClaw became my attempt to solve that problem without pretending the model should be in charge.
Planifier came from the same frustration. A trade idea should be more than a sentence and a target. It should become a plan somebody can understand, question, practice, and review after the fact.
CoinFox is the outward-facing part. I want people to be able to publish an idea, explain it well, be challenged, earn recognition, and keep a real record. A win should stay attached to the original reasoning. So should a mistake and the correction that followed it.
I am building all of this as a solo founder, outside of a full-time job, and I am learning a lot as I go. Some parts work. Some are still rough. I would rather be honest about that than act like the finished company already exists.
I do not know exactly how far I can take it. I know I would regret never finding out.
Brendan Paratore
07 / Current stage
Eastern Pine Ventures is early and actively building. Parts of FoxClaw, Planifier, and CoinFox already work in local or controlled settings. Other parts are still on the bench. The goal right now is a repeatable supervised workflow with strong evidence, clear authority, and no confusion about what has actually been proven.
The model is not an oracle. The system should make it easier to question.
08 / Contact
If this work interests you, if you see something we should study, or if you simply want to talk markets and systems, send me a note.