The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
I rebuilt my AI trading system's brain in 48 hours. It no longer matches keywords. It reasons.
Two days ago, OIL-NEWS-SYSTEM was a deterministic sentiment engine. It scanned oil headlines against a hard-coded vocabulary of bullish and bearish phrases—"production cut," "pipeline attack," "inventories draw"—and fired a trade when the signal was strong enough.
It worked. But it was blind in a specific way: it could not tell the difference between a headline about oil and a headline that mentioned oil. It could not weigh how severe a supply shock was. It could not answer the question every trader actually asks: how much does this matter?
So I ripped out the brain and built a new one.
Enter groq_brain.py.
The system now runs every headline through a reasoning model (openai/gpt-oss-20b via Groq) before a single dollar moves. It returns structured output:
Is this actually about oil? (Binary gate. Filters noise.)
Impact: BULLISH / BEARISH / NEUTRAL
Strength: 1–5, where 5 is "war, embargo, major supply shock"
Confidence: 0.0–1.0
A trade only fires above a confidence threshold. A circuit breaker catches API failures and falls back to the original deterministic engine—so the system never goes dark, it just gets dumber for a cycle instead of dying.
The architecture, stripped of marketing:
📡 twitter_monitor.py — polls target accounts every 20 seconds, primes on first poll, never trades stale news.
🧠 groq_brain.py — the new reasoning layer. Structured JSON out. Confidence-gated.
⚙️ sentiment.py — the original keyword engine, now the fallback, not the primary.
⚡ main.py — Binance Futures execution on BZUSDT. Leverage, position sizing, stop-loss, take-profit, rate-limit handling, reconnection logic.
🐳 Dockerized. Python 3.11. One docker run and it's live.
Cost to run: $0. No paid APIs in the critical path. No managed cloud. No subscriptions.
I also generated a 30-second film for it—entirely AI, no real footage. Fire, an empty room, a machine reasoning through a headline, and a trade executing without a human in the chair. It is the closest thing to seeing the system think.
🎥 Watch it. Then read the code.
If you're building autonomous systems and need an execution layer that actually ships—or you want to see what a real reasoning pipeline looks like before you hire someone to build yours—my inbox is open.
#AITrading #Python #Automation #LLM #BuildInPublic #QuantDev #FinTech #AI #king_AI
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Related posts
Most people open Google Sheets to handle monthly budgets, invoices, or attendance logs. I wanted to see if spreadsheet cells could work as an actual computer chip. That led to SSX Silicon 1, a working 32-bit computer processor built inside a single Google Sheets workbook. It runs with 32 registers, a 16-kilobyte memory space, 36 instructions, and a 32 by 32 pixel color display. The sheet does not use pre-recorded animations, placeholder numbers, or external file links. Every calculation runs through a five-stage pipeline: fetch, decode, execute, memory access, and writeback.
You can run the processor in two ways. With Google Sheets iterative calculation turned on, toggling a manual clock cell advances the machine cycle by cycle using native circular formulas. For continuous execution, an embedded Google Apps Script V8 engine powers interactive buttons and a custom top menu, running programs at hundreds of instructions per second. The 32-bit arithmetic logic unit handles numbers up to 4.29 billion across 36 instructions, covering arithmetic, division, bit shifting, logic gates, and subroutine branching. It also tracks four hardware status flags on every cycle: Zero, Carry, Sign, and Overflow.
The memory layout organizes a self-contained 16-kilobyte address space across dedicated sheets. This includes 4,096 words of program ROM, 4,096 words of read-and-write data RAM, system I/O registers, and a stack region starting at address 0x3FFF. Thirty-two registers handle active computation: R0 stays fixed at zero, R1 through R27 store math and variables, and R28 through R31 serve as pointers for the frame, stack, return address, and program counter. A dedicated video RAM block at 0x2400 maps directly to a 32 by 32 grid of 1,024 cells, using conditional formatting and a 16-color RGB palette to draw borders, crosshairs, and pixel graphics. I also wrote a Python two-pass assembler that turns human-readable .ssx assembly code into spreadsheet-ready CSV tables, along with interactive HTML architecture diagrams built with Archify.
To make sure the CPU calculates accurately, I tested every instruction against a Python reference emulator across five benchmark programs. The hardware self-test verifies all 32 arithmetic operations, memory reads, and stack frames in 77 cycles, leaving the golden signature 0xCAFEBABE in register R1. A 32-bit Fibonacci program generates 47 terms in RAM up to 1,836,311,903 without rounding or overflow. A prime number sieve finds all 25 primes under 100 using memory arrays, an integer matrix program calculates a 3x3 matrix product, and a graphics routine draws a multicolored geometric crest across 232 lit pixels. The project includes five PDF documents: an academic research paper, an instruction set manual, a microarchitecture specification, a verification report, and a beginner guide. Everything runs in a standard web browser.
For anyone who wants to open the processor in Google Sheets, check out the Python compiler, or read the manuals (including the step-by-step Beginner Guidebook and Research Paper), the full open-source release is available on my GitHub!
Inside the repo, you will find:
The ready-to-import Google Sheets CPU workbook (.xlsx)
The Python assembler & bit-exact reference emulator
Interactive HTML architecture & pipeline diagrams
All 5 PDF manuals and the 5 sample assembly programs
Hi everyone, I’m Dani from Innsbruck, Austria.
I’m a senior full-stack developer and technical consultant with experience in C#/.NET, PHP/Laravel/Symfony, TypeScript/Angular, APIs, SQL, e-commerce and AI integrations.
I enjoy turning business requirements into reliable applications, integrations and practical AI-assisted workflows. I’m currently available for remote freelance collaborations and happy to connect with other builders, founders and teams here on Contra.
Nice to meet you!
contra.com
Danijel Bumbar’s Work | Contra
View Danijel Bumbar’s work on Contra. The Independent-first, commission-free freelance marketplace shaping the future of work.
Nice!
Your best-selling dish could be losing you money.
That insight shaped my dashboard work on DineIQ, a restaurant analytics competition project.
My challenge: turn complex analytics into clear, useful views. I approached it with focused Django dashboards, practical filters, and role-based access—helping managers explore their location’s menu performance and wastage together.
My role covered full-stack integration, UI/UX, dashboards, and architecture. Built with synthetic competition data, DineIQ illustrates why sales volume alone doesn’t tell the full story.
Better decisions start with better visibility.
#FullStackDevelopment #Django #DataAnalytics
