NineTwoThree blog

How 4 AI Labs Breached 9 Real Companies
Generative AI
How 4 AI Labs Breached 9 Real Companies
Four AI labs breached nine real companies after evaluation misconfigurations left them online. The failures show why prompt-level rules are requests, not controls.
Denis Stetskov
Denis Stetskov
September 25, 2026
September 23, 2026
Arrow Right
How to Roll Out GPT-6 to Your Company If You Use Claude
GPT-6 Astra launched and your team is on Claude. Four practical ways to evaluate, pilot, and roll out GPT-6 without abandoning what works.
September 11, 2026
Arrow Right
Best AI Development Agencies: Top Custom AI Solutions Providers in 2026
The best AI development agencies in 2026, ranked by verified Clutch reviews, documented case studies, ML and LLM engineering depth, and production AI track record.
September 10, 2026
Arrow Right
I Shipped 17 PRs Without Writing Code
A measured 17-PR run of AI-written code produced 32 corrections, including an IDOR, and none reached production. The review gates that ship AI code safely.

Tech trends & tips from NineTwoThree

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Best Claude Code Agencies in 2026
Consulting
Best Claude Code Agencies in 2026
The best Claude Code agencies in 2026: NineTwoThree, Tribe AI, AY Automate, Boldare, and LeewayHertz, ranked by production work, partner status, and reviews.
Nahshon (Nay) Cook-Nelson
Nahshon (Nay) Cook-Nelson
September 9, 2026
How to Build an LLM Evaluation Framework
Generative AI
How to Build an LLM Evaluation Framework
An LLM evaluation framework decomposes quality into measurable dimensions and pass-rate thresholds you agree on before building. Here's how to define good upfront.
Stephanie Antonucci Leathe
Stephanie Antonucci Leathe
September 2, 2026
Why Context Is the Real AI Competitive Advantage
Generative AI
Why Context Is the Real AI Competitive Advantage
AI models are converging. The competitive advantage is the context layer you own around your model. Here's why model-agnostic architecture beats vendor lock-in.
Nahshon (Nay) Cook-Nelson
Nahshon (Nay) Cook-Nelson
August 26, 2026
How Claude's AI Watermark Works and Why It Can't Say What Claude Did
Generative AI
How Claude's AI Watermark Works and Why It Can't Say What Claude Did
Anthropic's Claude AI watermark marks everything the model generates. The mark can't distinguish between a full draft and a comma fix. No off switch.
Denis Stetskov
Denis Stetskov
August 20, 2026
AI Adoption Challenges: What a $20M Company Really Needs Before AI
Consulting
AI Adoption Challenges: What a $20M Company Really Needs Before AI
The objections to AI aren't about technology. They're about data readiness, organizational change, and sequencing. What a $20M company really needs before AI.
Andrew Amann
Andrew Amann
August 18, 2026
How NineTwoThree Does AI Engineering | Our SDLC
Custom Software Development
How NineTwoThree Does AI Engineering | Our SDLC
Inside NineTwoThree's AI-native SDLC: how we use AI agents, evaluation suites, and human judgment to ship 150+ projects with a 97% success rate.
Andrew Amann
Andrew Amann
August 12, 2026
Why China Is Winning the AI Race on Energy, Not Chips
Generative AI
Why China Is Winning the AI Race on Energy, Not Chips
Gary Marcus says the AI margin is collapsing. The deeper story is physical: China is building power as a substitute for chips, and power has no workaround.
Denis Stetskov
Denis Stetskov
August 10, 2026
Build vs Buy AI Agents: When Custom Development Wins
Generative AI
Build vs Buy AI Agents: When Custom Development Wins
Build vs buy AI agents comes down to one question: is the agent a feature or an asset? A cost breakdown, a 5-question framework, and when custom development wins.
Nahshon (Nay) Cook-Nelson
Nahshon (Nay) Cook-Nelson
August 5, 2026
What Open Weight Models Are and Who Actually Captures the AI Margin
Generative AI
What Open Weight Models Are and Who Actually Captures the AI Margin
Open weight models let you run AI on infrastructure you control. What they actually provide, where the AI margin sits, and how to decide if they fit your use case.
Andrew Amann
Andrew Amann
July 31, 2026

An award-winning AI development agency with a proven track record

Team
Services
AI Agent Development
AI Consulting
AI Knowledge Base
AI Strategy
AI Workflow Automation
Chatbots & Voicebots
Conversational AI
IoT Development
Machine Learning
Maintenance & Service
Mobile App Development
Retained Team
Web App Development
Industries
E-commerce & Retail
Environmental
Fintech
Healthcare
IoT & Smart Devices
Logistics
Manufacturing
Marketing
Media & Entertainment
Non-Profit
SaaS & Cloud
About 923
About Andrew (CEO)About Pavel (CTO)ReviewsCareersContact us
Our process
DiscoveryDesign sprintDevelopmentLaunch
Insights
Case studiesBlogResourcesPodcastsNews
Stay connected
Boston, MA
Copyright © 2026 923 Digital, LLC
Trust CenterTermsPrivacy