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
Skamira Forecast – Real-Time Monitoring & Alerting Platform
• Architected a 3-layer lakehouse (Bronze / Silver / Gold) on PostgreSQL with Airflow DAGs orchestrating realtime data ingestion and automated technical-indicator computation (RSI, MACD).
• Implemented a Kafka-based streaming pipeline for continuous event ingestion and anomaly/threshold alerting; containerized the full stack with Docker Compose for reproducible deployments.
• Trained an ensemble of 5 GRU models (TensorFlow/Keras) tracked in MLflow; achieved MAPE < 3% and R² > 0.95 on 7-day forecasts.
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
Trending
Claude
Claude has entered the design space. How are you using Claude Design?
Contra University
Learn from expert creatives how to earn more using next-gen AI tools.
creativeaiflow
Creative AI workflows are evolving. What tools do you use, and what are their strengths and weaknesses?
freelancerlife
Freelancer life is wins, pivots, and everything in between. What’s yours right now?
Related posts
Talon Audit — a privacy-first website exposure and remediation platform
I built Talon Audit to solve a problem most website tools handle badly: they detect issues, but rarely help teams understand what matters, what to fix first, or whether the fix actually worked.
Talon analyses public-facing websites across four connected domains:
Exposure — browser protections, cookies, scripts, TLS, and visible configuration risks
Integrity — broken resources, routes, errors, forms, and reliability regressions
Experience — accessibility, mobile usability, speed, and interaction friction
Conversion — messaging, trust signals, pricing paths, forms, and calls to action
The product follows a complete improvement workflow:
Detect → Explain → Fix → Verify → Monitor
Every finding is structured around evidence, severity, confidence, impact, remediation guidance, and a verification method.
I led the project from strategy to launch, including:
product positioning and methodology
UX/UI and design system
full-stack implementation
AI-assisted analysis and remediation workflows
privacy and security boundaries
pricing, demo experience, and launch
Talon performs passive, low-impact analysis only. It does not exploit systems or claim to replace authorised penetration testing.
Built through a terminal-first AI-native workflow using Codex 5.5 and Claude Fable.
Live product: talonaudit.com
Role: Product Strategy, Product Design, Full-Stack Development, AI Development
Structuring findings around evidence, severity, and verification methods makes remediation so much clearer. How are you handling false positives during passive analysis?
Just joined the @flowstep_ai Challenge.
Started exploring my idea and generated the first prototype. There's still a lot to improve, but it's exciting to see the concept taking shape.
Looking forward to sharing the progress.
#flowstepchallenge
Excited to see what it becomes!
Solo Architect & Developer: Engineered the end-to-end multi-agent AI infrastructure
Project description: I engineered a production multi-agent AI system acting as a virtual operations team. Using a hub-and-spoke architecture, 7 specialized Claude-powered agents coordinate through a PostgreSQL task queue to autonomously handle marketing, SEO, and coding tasks. The system is controlled entirely via Telegram and features a strict two-stage human-in-the-loop approval gate, ensuring no external actions occur without explicit sign-off. Built on Node.js and Render, I implemented robust security hardening, AES encryption, and multiple API integrations to deliver a secure, highly scalable AI platform.
