<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/the-economics-of-agent-scale-tokens-roi-and-building-platforms-for-ai-first-teams-part-2--1vqmchwpu" -->

---
title: The economics of agent scale: tokens, ROI, and building...
description: A podcast conversation between Stack Overflow&#x27;s Peter O&#x27;Connor and Google&#x27;s Andi Gutmans (Head of Agentic Data Cloud) explores the economics of running AI...
canonical: https://daily.dev/posts/the-economics-of-agent-scale-tokens-roi-and-building-platforms-for-ai-first-teams-part-2--1vqmchwpu
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: The economics of agent scale: tokens, ROI, and building platforms for AI-first teams (Part 2) | daily.dev
og:description: A podcast conversation between Stack Overflow&#x27;s Peter O&#x27;Connor and Google&#x27;s Andi Gutmans (Head of Agentic Data Cloud) explores the economics of running AI...
og:url: https://daily.dev/posts/the-economics-of-agent-scale-tokens-roi-and-building-platforms-for-ai-first-teams-part-2--1vqmchwpu
og:image: https://api.daily.dev/og/posts/1vqMcHWPU.png
og:image:alt: The economics of agent scale: tokens, ROI, and building platforms for AI-first teams (Part 2)
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# The economics of agent scale: tokens, ROI, and building platforms for AI-first teams (Part 2)

**[Stack Overflow Blog](https://daily.dev/sources/stackov)** · 24 min read · 0 upvotes · 0 comments

## Summary

A podcast conversation between Stack Overflow's Peter O'Connor and Google's Andi Gutmans (Head of Agentic Data Cloud) explores the economics of running AI agents at scale. Gutmans argues the model is rarely the bottleneck anymore; the harder problem is minimizing context to get reliable outcomes at the lowest cost, calling 'token maxing' the wrong goal. They discuss cost governance as agent counts scale into the dozens or hundreds per employee, Google's integrated approach of hill-climbing models and data together via DeepMind, and skepticism toward vendors claiming to have solved the context problem. They also cover building platform infrastructure to treat 'the agent' as a persona alongside developers and data scientists, referencing Spanner as a precedent for shifting correctness burden from developer to system. The episode closes with career advice for aspiring engineers: fundamentals, agentic fluency, and business judgment all matter.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://stackoverflow.blog/2026/09/03/the-economics-of-agent-scale>

## Questions this post answers

### Why does Andi Gutmans say token maxing is the wrong goal for AI agents?

Because the goal should be reaching differentiated outcomes and measurable ROI, not processing more tokens. Andi Gutmans, Head of Agentic Data Cloud at Google, argues the objective is finding the least amount of context needed for a reliable outcome at the lowest cost, using the least sophisticated model that gets the job done, since models today are often already good enough for many tasks.

_daily.dev surfaces practical takes on agent cost governance for teams tuning their own AI workflows._

### How did Google originally solve the scaling problems that MySQL had for YouTube?

YouTube initially ran on MySQL until it stopped scaling, prompting Google to build Vitess as a sharding layer, and eventually Spanner, a distributed database that shifts scalability ownership from the developer onto the data system itself. Andi Gutmans cites this as the model for how Google now wants to build infrastructure that similarly removes scaling burden from agents.

_engineers weighing database scaling strategies can find more infrastructure deep dives on daily.dev._

### What career advice does Andi Gutmans give to someone starting a computer science degree today given the rise of AI agents?

Fundamentals like math and systems architecture still matter a lot, agentic coding fluency is now essential, and understanding business problems is increasingly valuable, comparable to how an MBA once supplemented a CS degree. Andi Gutmans, Head of Agentic Data Cloud at Google, says employers will look for people who can drive outsized impact by combining these three areas.

_students shaping a CS learning path can track how agentic skills are reshaping careers on daily.dev._

## Similar posts on daily.dev

- [When code is abundant](https://daily.dev/posts/when-code-is-abundant-gezbqtizx) · GitLab · 0 upvotes · 0 comments
- [How enterprise leaders are scaling AI agents across their organization](https://daily.dev/posts/how-enterprise-leaders-are-scaling-ai-agents-across-their-organization-gaahtnpqz) · databricks · 0 upvotes · 0 comments

---

Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#gcp](https://daily.dev/tags/gcp), [#observability](https://daily.dev/tags/observability), [#platform-engineering](https://daily.dev/tags/platform-engineering)

[View this post on daily.dev](https://daily.dev/posts/the-economics-of-agent-scale-tokens-roi-and-building-platforms-for-ai-first-teams-part-2--1vqmchwpu)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"The economics of agent scale: tokens, ROI, and building platforms for AI-first teams (Part 2)","url":"https://daily.dev/posts/the-economics-of-agent-scale-tokens-roi-and-building-platforms-for-ai-first-teams-part-2--1vqmchwpu","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-economics-of-agent-scale-tokens-roi-and-building-platforms-for-ai-first-teams-part-2--1vqmchwpu"},"datePublished":"2026-09-03T04:34:51.573Z","dateModified":"2026-09-03T08:34:05.561Z","description":"A podcast conversation between Stack Overflow's Peter O'Connor and Google's Andi Gutmans (Head of Agentic Data Cloud) explores the economics of running AI...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/bd7543439f0b987e9af854c2a07f33e4?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/bd7543439f0b987e9af854c2a07f33e4?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Stack Overflow Blog","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Stack Overflow Blog","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/003e14873238469caea60ebdd34508ae","url":"https://daily.dev/sources/stackov"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/the-economics-of-agent-scale-tokens-roi-and-building-platforms-for-ai-first-teams-part-2--1vqmchwpu","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,gcp,observability,platform-engineering","timeRequired":"PT24M"}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Stack Overflow Blog","item":"https://daily.dev/sources/stackov"},{"@type":"ListItem","position":3,"name":"The economics of agent scale: tokens, ROI, and building platforms for AI-first teams (Part 2)"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-economics-of-agent-scale-tokens-roi-and-building-platforms-for-ai-first-teams-part-2--1vqmchwpu#faq","mainEntity":[{"@type":"Question","name":"Why does Andi Gutmans say token maxing is the wrong goal for AI agents?","acceptedAnswer":{"@type":"Answer","text":"Because the goal should be reaching differentiated outcomes and measurable ROI, not processing more tokens. Andi Gutmans, Head of Agentic Data Cloud at Google, argues the objective is finding the least amount of context needed for a reliable outcome at the lowest cost, using the least sophisticated model that gets the job done, since models today are often already good enough for many tasks. daily.dev surfaces practical takes on agent cost governance for teams tuning their own AI workflows."}},{"@type":"Question","name":"How did Google originally solve the scaling problems that MySQL had for YouTube?","acceptedAnswer":{"@type":"Answer","text":"YouTube initially ran on MySQL until it stopped scaling, prompting Google to build Vitess as a sharding layer, and eventually Spanner, a distributed database that shifts scalability ownership from the developer onto the data system itself. Andi Gutmans cites this as the model for how Google now wants to build infrastructure that similarly removes scaling burden from agents. engineers weighing database scaling strategies can find more infrastructure deep dives on daily.dev."}},{"@type":"Question","name":"What career advice does Andi Gutmans give to someone starting a computer science degree today given the rise of AI agents?","acceptedAnswer":{"@type":"Answer","text":"Fundamentals like math and systems architecture still matter a lot, agentic coding fluency is now essential, and understanding business problems is increasingly valuable, comparable to how an MBA once supplemented a CS degree. Andi Gutmans, Head of Agentic Data Cloud at Google, says employers will look for people who can drive outsized impact by combining these three areas. students shaping a CS learning path can track how agentic skills are reshaping careers on daily.dev."}}]}
```

