Agentic Workforce Company
Ops & Experiments · Started, Then Abandoned

I built a nine-feed trading pipeline — then shut it down.

A multi-agent crypto and prediction-market signal system: nine data feeds, a convergence scorer, whale-alert and contrarian channels, autonomous paper-trading. I measured a losing edge honestly — 81 trades, ~26% win rate, -$30.55 — killed the over-engineered version, tried a leaner one, and wound the whole thing down. Zero real dollars risked.

AbandonedROLE Owner-OperatorTIMEFRAME Feb 13 – Mar 22, 2026CAPITAL Paper only · $0 real
CategoryOps & experiments
System9-feed multi-agent pipeline
Result-$30.55 · ~26% win · paper
Risk posturePaper-traded · zero real capital
StatusAbandoned · the judgment to stop
The Problem

Could a fleet of agents find a real edge?

The fleet’s stated objective was “make money.” Markets are the obvious place to test it.

The hypothesis, borrowed from a third-party course claiming an “AI contrarian” strategy, was that agents could bet against the crowd.

Prediction and crypto markets are where you’d test whether AI’s real advantage — reading and synthesizing huge amounts of data faster than a human — translates into a tradeable edge.

The whole point of the experiment was to find out if that edge actually existed for us, on live markets, with our own pipeline — before ever committing real capital.

Set the Record Straight

Whose numbers are whose.

Two very different figures get confused here. This page keeps them apart on purpose.

▲ The course’s claim · NOT mine

“~11,000% ROI” belonged to the course.

  • The eye-catching ~11,000% ROI figure came from the third-party course that inspired the experiment.
  • It was a marketing claim about someone else’s strategy — never a result my system produced.

Do not read the course’s advertised return as this project’s outcome. It is shown only to explain where the hypothesis came from.

● My measured result

A small net loss on paper.

  • My system’s actual, measured outcome: -$30.55 across 81 closed paper trades.
  • 25.93% win rate. A losing edge, measured honestly.
  • Zero real dollars were ever at risk — paper-traded throughout.
The Approach

Build it properly, then audit it honestly.

V1 was the maximalist version. The decisive move was the honest audit that followed.

V1 was a nine-feed convergence pipeline feeding a scorer, with autonomous paper-traders auto-executing on high-convergence signals, whale-tracking, and a dashboard. It ran, it traded on paper, and it lost — and its intelligence layer degraded to garbage, with 5 of 9 feeds failing and signals coming back null.

Rather than tweak endlessly, I had the agent lay out the full profit-and-loss by category. So I killed V1 entirely — deleted the crons, purged 210 files — then tried a leaner research-only V2, and eventually stood that down too.

Over-engineered and under-performing — complex architecture producing zero intelligence. So I killed it, rather than polish a losing system.

— The kill diagnosis · from the audit
Architecture & Stack

The V1 pipeline.

Nine feeds converge on a scorer; only multi-source agreement flags a trade. High-convergence signals auto-execute on paper-traders; whale and contrarian signals fan out to alert channels; everything syncs to a dashboard.

Fig.01 — Trading-Signal Pipeline (V1)Multi-agent trading-signal pipelineNine data feeds on the left flow into a convergence scorer in the center, which requires multiple independent feeds to agree before flagging a trade. High-convergence signals route to autonomous paper-traders, which execute paper-only positions. The scorer also fans whale and contrarian signals out to alert channels, and all activity syncs to a dashboard backed by Postgres and Convex. Paper only, no real capital.9 DATA FEEDSFear & GreedCoinGecko divergenceWhale-alert flowsManifold arbitrageNews sentimentLunarCrush socialReddit sentimentTwitter sentimentOn-chain analyticsSIGNAL LAYERConvergencescorer≥70 tradeable · ≥85 autoEXECUTIONPaper-tradersno real capitalALERT CHANNELSwhale · contrariannear-empty in practiceDATA / INFRADashboardPostgres · ConvexBACKTESTERreplay vs rules

Nine feeds fan into a convergence scorer that only flags a trade when multiple independent sources agree. High-convergence signals route to paper-traders no real capital — while whale and contrarian signals fan out to alert channels. Everything syncs to a dashboard on Postgres and Convex, with a backtester replaying trades against new rules. On the final run, 5 of 9 feeds were failing and the scorer produced zero tradeable signals.

Agents

Master ControlRazor (technical)Echo (research)specialist sub-agents

Data Feeds (9)

Fear & GreedCoinGeckowhale-alertManifold arbitragenews sentimentLunarCrushRedditTwitteron-chain

Signal Layer

convergence scorercontrarian detectorwhale-tracking

Execution

paper-traders (crypto)paper Polymarketposition managementpaper only · no real capital

Data / Infra

Neon PostgreSQLConvex syncbacktester

I/O

alert channelstrading groupMission Control dashboard
By the Numbers

The audit, unvarnished.

Feb–Mar 2026 · verified from the honest P&L audit · paper trades throughout.

Closed trades
81
all paper-traded · zero real dollars
Win rate
25.93%
a losing edge, measured honestly
Total P&L
-$30.55
small net loss · paper only
V1 files purged
210
628 → 418 files · dead code removed
5 of 9
data feeds failing on the last run — 34 signals returned all-neutral, all-null
129 → 40
Python scripts after the purge — 89 removed as over-engineering
$0
real capital exposed at any point — paper-traded from start to finish
Category (paper)RecordP&L
Crypto Up/Down (lucky binary bets)a few big wins+$1,991
Elon-tweet markets0 wins / 10 lossesnet negative
Crypto Spotnet negative-$1,882
Total · 81 trades · 25.93% win ratemeasured-$30.55

The only positive category was essentially coin-flip binary markets where a few big bets got lucky; everything else was flat or negative. The infrastructure was built out well beyond what it ever produced — the alert channels carried a handful of messages over about a day, then went quiet.

The Verdict

The valuable outcome was the decision to stop.

I built a genuinely sophisticated system, ran it on paper, and it did not find an edge. The portfolio value isn’t a return — it’s the discipline.

● Decision discipline

Measured it. Called it. Cut it.

  • Forced a real, category-level P&L audit instead of more feature work.
  • Accepted a losing verdict on a system I’d invested real build effort in.
  • Killed and purged the over-engineered V1 — 210 files removed so it couldn’t rot in the workspace.
  • Tried a leaner V2 hypothesis (research-driven, conviction-sized, logged reasoning).
  • Wound V2 down too when it didn’t earn its keep — crons removed, trading group cleared Mar 22.
▲ Status · abandoned experiment

Framed exactly as what it was.

  • This is a started-then-abandoned experiment — not a profit, not a running system.
  • It shows range: multi-agent architecture, market-data engineering, backtesting, dashboards.
  • More importantly, it shows the judgment to stop — most people keep polishing a losing system.

No real capital was ever exposed. Do not imply this made money or is still running. Its value is breadth plus the discipline to walk away.

Kill & Purge

From maximalist to wound down.

Workspace file count across the experiment
V1 · maximalist628files · 9 feeds · auto-execute
Purged418210 removed · dead code cut
Lean V2 → wound downStoppedresearch-only · then shut down
V1 → V2 → Wind-Down

How the experiment ended.

Feb 13 · Build
Maximalist V1 stood up

Nine-feed convergence pipeline, autonomous paper-traders auto-executing on high-convergence signals, whale-tracking, and a synced dashboard.

Run · Degrade
Intelligence layer decayed

It traded on paper and lost; 5 of 9 feeds failed and the scorer produced 34 neutral, null signals — zero tradeable output.

Audit · Kill
Honest P&L, then the kill call

A category-level audit exposed the losing edge (-$30.55, ~26%). Diagnosis: over-engineered and under-performing. V1 crons deleted, 210 files purged.

V2 · Lean
Research-only rebuild

Inverted the thesis: no signal-spam, one $10K paper account, conviction sizing, max 3 trades/day, capture the “why” behind every trade.

Mar 22 · Stop
Whole experiment wound down

V2 research crons removed on strategy shutdown; the trading group was cleared out. No real capital was ever exposed.

My Role

I owned the hypothesis, the risk controls, and the kill.

I scoped the experiment, directed the build, set the safety posture — paper-trading only, no real money ever went in — and forced the honest audit.

When performance was ambiguous I demanded the category-level P&L rather than more feature work, made the call to kill V1 and permanently purge it, defined the leaner V2 thesis, and decided to wind the whole thing down. The agents built and traded; I owned the hypothesis, the risk controls, the audit, and the kill decision.

Skills Demonstrated

What this took.

Multi-agent system designMarket-data pipeline engineering (9-feed ingestion)Convergence scoringBacktesting & honest performance measurementRisk discipline (paper-only, no real capital)Knowing when to kill a projectHypothesis iteration (maximalist → lean)Prediction-market / crypto domain modelingPostgres / Convex data plumbingDashboard + alerting design

Want an operator who measures honestly and knows when to stop?