
Overview
# HackGlobal 2026 - Super Secure Syst v23
Problem: Single gas detectors only monitor ONE gas, no action. Food, room, skin, dye, forensic and gem checks are done separately and slowly.
Solution: One agent with 6 Detect tools in one dashboard.
6 Detect:
1. Food - RGB IMX219 + NDIR 4.26um CO2
2. Room / Bio Healthy Room - SHT41 + NDIR + Thermal: dust, humidity, dirty area at risk for mold, pests, bacteria. Status: Clean / Needs Cleaning / Risk. Monitoring + prevention prototype, not a medical sterilization claim.
3. Derma - Multispectral skin check preview
4. Naphtol - NDIR 3.4um Naphtol + 3.3um H2O/OH
5. Forensic Trace - Trace detection preview
6. Gem Authenticity - 7.7um emissivity + RGB
How it works:
[NDIR 4-CH + IMX219 + MLX90640] -> [ESP32-S3 + Pump] -> [MQTT TLS] -> [Vultr Compute + Vultr Serverless Inference] -> [Streamlit Observatory + ESC/POS Certificate]
Safety: 30-Second Safety Observatory. Agent converts "CO 40ppm" to "Open damper 30%, evac zone 2, notify tech".
Vultr Setup: Backend on Vultr Compute Singapore, close to Jakarta for low latency. Inference via Vultr Serverless Inference. Repo public and runnable: git clone https://github.com/AIchaliveret/super-secure-syst-v23.git, pip install -r requirements.txt, streamlit run app.py
Honest trade-off: Live demo uses simulated sensor data so judges can test without hardware. Real hardware uses ESP32 GPIO16/17 UART2 to NDIR, GPIO21/22 I2C to O2/H2S, GPIO27 PWM pump.
Supporting repos: super-infrared (IR core) + ir-7detik-lensa-infra (fast 7-sec IR lens)
How it uses Vultr
Backend runs on Vultr Compute in Singapore, chosen because it is closest to Jakarta for low latency. Flow: ESP32-S3 (NDIR 4-CH + IMX219 + MLX90640 + SHT41) -> MQTT TLS -> Vultr Compute -> Vultr Serverless Inference (Llama) -> Streamlit Observatory Dashboard. Vultr Serverless Inference does the 6 Detect reasoning: Food, Bio Healthy Room, Derma, Naphtol, Forensic Trace, Gem Authenticity. It converts raw sensor values into actions, e.g. CO 40ppm -> Open damper 30%, evac zone 2, notify tech. Vultr Compute hosts the Streamlit app, MQTT broker, RAG Vector Search for safety rules, and ESC/POS certificate generator. Wiring: NDIR via UART2 GPIO16/17, O2/H2S via I2C GPIO21/22, Pump via GPIO27 PWM. 30-Second Safety Observatory + 7-Second IR Lens for fast check. Bio Healthy Room uses the same pipeline: dust + humidity + dirty area -> Clean / Needs Cleaning / Risk. Monitoring + prevention, not medical sterilization. Live demo uses simulated sensor data so judges can test without hardware. Real hardware uses the same MQTT JSON format.
No comments yet
Be the first to comment on this project.

FloatGPT — Omnipresent, Local-First AI Desktop Companion
FloatGPT is an omnipresent desktop AI companion that floats over Windows to automate OS workflows, plan tasks, and generate real-time executive dossiers. Operating 100% privately on your PC or powered by the cloud, it transforms daily computing with screen awareness and calm execution.

ClaimTrace
ClaimTrace audits the citations in a scientific manuscript: it checks that every reference exists and is described correctly, and that each cited source actually says what the citing sentence claims, with a verbatim quote as evidence. On 1,728 reference test cases it never marked a fabricated or altered reference as verified.

SeeHere: audio descriptions for lecture videos, grounded in visual evidence
When a lecturer says “look at this chart”, blind and low-vision students hear nothing. SeeHere watches a recorded lecture, finds what is on screen but never said aloud, and writes short audio descriptions that a CP-SAT solver fits into natural pauses. Every line is linked to its evidence (keyframe, OCR, transcript) and checked, and uncertain claims are held for an instructor. Live on Vultr Cloud Compute with Vultr Serverless Inference.


