CURRENT | Foundations

Field Guide

Three Territories of Web Automation Reliability

By Rina Takahashi— January 28, 2026

Feature image for article: Three Territories of Web Automation Reliability

Your automation runs flawlessly through a hundred test cycles—clean data, zero errors, perfect repeatability. Scale to production and three weeks later you're explaining why revenue dashboards went blank overnight. The script still works. Something else broke. At TinyFish, we see this constantly: teams prototype in one operational reality, then deploy into another. By the time they realize the mismatch, downstream systems already depend on infrastructure that can't deliver what production actually demands.

Field Guide

Three Territories of Web Automation Reliability

By Rina Takahashi— January 28, 2026

Your automation runs flawlessly through a hundred test cycles—clean data, zero errors, perfect repeatability. Scale to production and three weeks later you're explaining why revenue dashboards went blank overnight. The script still works. Something else broke. At TinyFish, we see this constantly: teams prototype in one operational reality, then deploy into another. By the time they realize the mismatch, downstream systems already depend on infrastructure that can't deliver what production actually demands.

Tools & Techniques

Tools in Context

Selective Preservation Works When You Know What Matters Before You Look

Extract exactly what you need: the price, the availability flag, the product title. Nothing more. This approach assumes you can predict what matters before you look—that sites behave consistently enough to target specific data points without preserving surrounding context. For some sites, that assumption holds. Production reveals which ones, and the answer surfaces only after you've run thousands of extractions across months of operation.

Tools in Context

Comprehensive Preservation for Context You Can't Predict

Preserve everything first. Settled HTML after JavaScript execution, screenshots showing visual layout, network traces capturing the full request chain. Extract your data from that preserved state, then keep the state itself. Storage costs multiply, but teams choose this because the web surprises you—authentication flows that shift by region, bot detection that serves different content, layout changes that break extraction logic you can't debug without evidence.

Tools & Techniques

Tools in Context

Selective Preservation Works When You Know What Matters Before You Look

Extract exactly what you need: the price, the availability flag, the product title. Nothing more. This approach assumes you can predict what matters before you look—that sites behave consistently enough to target specific data points without preserving surrounding context. For some sites, that assumption holds. Production reveals which ones, and the answer surfaces only after you've run thousands of extractions across months of operation.

Tools in Context

Comprehensive Preservation for Context You Can't Predict

Preserve everything first. Settled HTML after JavaScript execution, screenshots showing visual layout, network traces capturing the full request chain. Extract your data from that preserved state, then keep the state itself. Storage costs multiply, but teams choose this because the web surprises you—authentication flows that shift by region, bot detection that serves different content, layout changes that break extraction logic you can't debug without evidence.

Tools & Techniques

Tools in Context

Selective Preservation Works When You Know What Matters Before You Look

Extract exactly what you need: the price, the availability flag, the product title. Nothing more. This approach assumes you can predict what matters before you look—that sites behave consistently enough to target specific data points without preserving surrounding context. For some sites, that assumption holds. Production reveals which ones, and the answer surfaces only after you've run thousands of extractions across months of operation.

Tools in Context

Comprehensive Preservation for Context You Can't Predict

Preserve everything first. Settled HTML after JavaScript execution, screenshots showing visual layout, network traces capturing the full request chain. Extract your data from that preserved state, then keep the state itself. Storage costs multiply, but teams choose this because the web surprises you—authentication flows that shift by region, bot detection that serves different content, layout changes that break extraction logic you can't debug without evidence.

In Dialogue With Complexity

An Interview With the Context Window on Why It Was Never Meant to Be a Security Boundary

In Dialogue With Complexity

An Interview With the Context Window on Why It Was Never Meant to Be a Security Boundary

Pattern Recognition

Companies Redesign Work Around AI Agents as Teammates

Organizations are restructuring hiring and roles to accommodate AI agents as operational colleagues. Entry-level hiring alterations jumped from 18% to 64% in one quarter. HCLTech trained 116,000 employees while creating orchestration engineers who shape agent execution and responsible AI engineers who build guardrails.

Companies expect 42% of their tech workforce to remain human by 2027, only a five-point drop despite rapid adoption. The work shifts from specialized execution to generalized oversight. Finance professionals focus on revenue growth and vendor negotiations while agents process invoices and reconciliations.

Successful deployment requires workforce redesign alongside technology. Gartner predicts 40% of projects will fail from poor integration and unclear ownership, not technical problems.

Pattern Recognition

Companies Redesign Work Around AI Agents as Teammates

Organizations are restructuring hiring and roles to accommodate AI agents as operational colleagues. Entry-level hiring alterations jumped from 18% to 64% in one quarter. HCLTech trained 116,000 employees while creating orchestration engineers who shape agent execution and responsible AI engineers who build guardrails.

Companies expect 42% of their tech workforce to remain human by 2027, only a five-point drop despite rapid adoption. The work shifts from specialized execution to generalized oversight. Finance professionals focus on revenue growth and vendor negotiations while agents process invoices and reconciliations.

Successful deployment requires workforce redesign alongside technology. Gartner predicts 40% of projects will fail from poor integration and unclear ownership, not technical problems.

Hiring acceleration:

Entry-level hiring changes jumped from 18% to 64% in one quarter, signaling rapid organizational transformation beyond technology.

Training scale:

HCLTech trained 116,000 employees in generative AI while creating new roles like orchestration and responsible AI engineers.

Workforce retention:

High-performing companies plan 50% permanent human staff by 2027 versus 42% average, valuing continued human expertise.

Efficiency gains:

KPMG's AI Career Companion achieved 99.75% time reduction generating skills architectures for 15,000 employees across the organization.

Essential skill:

92% of organizations expect managing AI agents as team members requiring oversight will become critical within five years.

Questions Worth Asking

Most teams ask the wrong questions before deploying agents. They focus on capability demonstrations instead of operational reality. The gap shows up fast.

We've watched this pattern repeat: frameworks get stripped out, integration takes three times longer than building, maintenance compounds until someone gets paged at 2am. The questions that actually predict success come from teams who've operated agents at scale and learned what breaks.

These six questions cut through vendor promises to what matters in production. They're born from observable failures: the infinite loops, the drift nobody noticed, the SLAs you can't actually meet. Ask them before you deploy, not after.

Questions Worth Asking

Most teams ask the wrong questions before deploying agents. They focus on capability demonstrations instead of operational reality. The gap shows up fast.

We've watched this pattern repeat: frameworks get stripped out, integration takes three times longer than building, maintenance compounds until someone gets paged at 2am. The questions that actually predict success come from teams who've operated agents at scale and learned what breaks.

These six questions cut through vendor promises to what matters in production. They're born from observable failures: the infinite loops, the drift nobody noticed, the SLAs you can't actually meet. Ask them before you deploy, not after.

Framework Dependency

Will This Become Invisible Infrastructure?

Production teams eventually strip out frameworks. The pattern repeats: what ships demos creates maintenance burden at scale. You need infrastructure that fades into the background, not another system demanding constant attention and specialized knowledge to operate.

Deployment Reality

How Wide Is Your Demo Gap?

Controlled environments hide what breaks under load. Agents enter infinite loops, produce inconsistent outputs, fail catastrophically when traffic patterns shift. The distance between "works on my machine" and "handles Tuesday afternoon reliably" predicts your first month in production.

SLA Feasibility

Can You Write Contracts on This?

Traditional SLAs assume deterministic behavior. Your agent chooses different approaches for identical problems. You're either measuring adherence after failures happen or maintaining guarantees in real time. Probabilistic systems struggle with contractual commitments for a reason.

Integration Overhead

Does Integration Dwarf Development Time?

Hooking agents into existing systems consistently takes longer than building them. Legacy infrastructure lacks modern APIs. Production agents live inside complex ecosystems with databases, tools, and workflows that weren't designed for probabilistic behavior. Budget accordingly.

Observability Requirements

Can You Debug Without Full Traces?

You can't fix what you can't see. Production issues require context: what the agent was thinking, which tools it called, what errors occurred. Systems that don't log full interaction traces, reasoning chains, and performance metrics leave you guessing during outages.

Maintenance Model

Who Owns This When It Drifts?

Evaluation suites need ongoing attention and clear ownership. Models drift as real-world data deviates from training distributions. Set-it-and-forget-it reliability doesn't exist in production. You're committing to continuous maintenance whether you planned for it or not.