#AI agents

#AI agents

OpenAI is introducing what it is calling, naturally, "stateful AI." The company has announced that it will soon offer a stateful runtime environment in partnership with Amazon, built to simplify the process of getting AI agents into production. It will run natively on Amazon Bedrock, be tailored for agentic workflows, and optimized for AWS infrastructure.

Running AI infrastructure costs are astronomical. Back in 2023, it was estimated that OpenAI spends around $700,000 per day to run ChatGPT—about 36 cents per query. However, in 2024 with the release of its higher-performing o3 model, some queries cost over $1,000 of computing power. Consequently, OpenAI CEO Sam Altman reports the company is even losing money on its $200 ChatGPT Pro subscriptions.

An AI agent is simply a model that receives input, follows defined goals and rules, makes step-by-step decisions, and uses tools to take actions. Instead of viewing AI agents as autonomous digital workers, break it down. First principles thinking says this definition captures the essence of how AI agents function and operate within business workflows.

The new battleground in banking is intelligent operations and scalable execution. In 2026, banking is about moving money smarter, faster, and with fewer humans in the middle. Across corporate finance and global retail operations, banks are experimenting with technology and operational design in ways that challenge long-held assumptions about scale, speed, and control. Three recent developments exemplify what's happening in money movement: Goldman Sachs deploying AI agents, Truist automating corporate receivables, and Nubank expanding abroad with a lean digital model.

Just consider a typical day in the life of a modern human: you glance at your phone while waiting for coffee to brew, skim headlines while half-listening to a podcast, mentally rehearse a client pitch while walking your child to school, reply "noted" on Slack during a meeting while updating a slide deck, check your bank balance while standing in line,

I'll run out of money.

We began GitHub Agentic Workflows as an investigation into a simple question: what does repository automation with strong guardrails look like in the era of AI coding agents? A natural place to start was GitHub Actions, the heart of scalable repository automation on GitHub. GitHub Agentic Workflows leverage LLMs' natural language understanding to let developers define automation goals in simple Markdown files describing the desired outcome.

Agentic AI workflows sit at the intersection of automation and decision-making. Unlike a standard workflow, where data flows through pre-defined steps, an agentic workflow gives a language model discretion. The model can decide when to act, when to pause, and when to invoke tools like web search, databases, or internal APIs. That flexibility is powerful - but also costly, fragile, and easy to misuse.