--- title: "Practitioner's Corner" canonical: "https://current.tinyfish.ai/issue/44/practitioners-corner" publication: "TinyFish" publisher: "TinyFish" language: "en-US" guide: "https://current.tinyfish.ai/llms.txt" permalink: "https://current.tinyfish.ai/issue/44/practitioners-corner" issue: 44 section: "practitioners-corner" description: "Lessons from the field—what we see building at scale" authors: ["Sevda Polat","Sable Whitford"] published: 2026-08-31 updated: 2026-08-28 articles: 14 --- # TinyFish > CURRENT is a journal about AI agents in production, for the engineering and product leaders who have to make them work. Drawing on experience building enterprise web agents, it explains why the demo succeeded and the deployment didn't, and what that means for choosing vendors, briefing the board, and avoiding mistakes others have already made. [Read on the website]() ## Issue 44 — Practitioner's Corner Lessons from the field—what we see building at scale ### Residual Case Composition - [The Queue After the Queue]() — By [Sevda Polat]() · Published 2026-08-28 · Format: Residual Case Composition When a telecom deployed a conversational agent, daily call volume to human agents barely moved. But the average call got longer. The system was resolving cases and making the human operation slower at the same time. Automating routine work strips out the predictable middle and leaves behind a concentrated residue of ambiguity, judgment calls, and emotional difficulty. Organizations celebrating an 80% automation rate may be running a human operation that works less well than the one it replaced. ### Builder Profile - [Alexandre Drouin and the Benchmark as a Theory of Work]() — By [Sevda Polat]() · Published 2026-08-28 · Format: Builder Profile Alexandre Drouin's WorkArena benchmark tests AI agents on enterprise tasks inside live ServiceNow software: filtering incident lists, ordering configured laptops, reading dashboard values. Every included task shares one property — it ends in a system state that a program can verify. That criterion is also a claim about work. Tasks whose outcomes live in database fields can be benchmarked; tasks whose value depends on interpretation, conversation, or judgment that never resolves into a checkable field cannot. Drouin's design choices show where that line falls more clearly than his scores do. ### Sorting as Primary Work - [A Triage Nurse on the Sorting Decision Nobody Measures]() — Published 2026-08-28 *Dolores "Dee" Gatewood is a composite — a fictional triage nurse assembled from published research, clinical literature, and the occupational knowledge that real triage nurses have described in peer-reviewed focus groups and grounded theory studies. She is not real. Her observations are.* ### The Emotional Residual - [Present but Pulling Back]() — By [Sevda Polat]() · Published 2026-08-28 A [field experiment on Alibaba's Taobao](https://arxiv.org/abs/2605.14830) tested what happens when an AI agent handles customer service and routes hard cases to humans. With 647 workers and 680,000 chats, the results split cleanly. Technical escalations — cases the agent couldn't resolve — preserved service quality. Emotional escalations — already-frustrated customers — produced 40% longer resolution times, ratings down nearly a full point, and recontact rates up six percentage points. ### External Reading - **What Happens to Workers When Agents Escalate Too Late** — [Source]() — Published 2026-08-28 Alibaba field experiment finds worker disengagement spikes when automated agents delay handoff of emotionally charged conversations. - **How 86 Deployed Agent Systems Actually Run in Production** — [Source]() — Published 2026-08-28 ICML study finds most production agents execute fewer than 10 steps and skip external frameworks entirely. - **Tool-Using Agents Mishandle Sensitive Data During Routine Tasks** — [Source]() — Published 2026-08-28 Joint safety evaluation shows agents expose private information in normal operations, without any adversarial prompting. - **When Workers Keep Discretion, the Failure Modes Change** — [Source]() — Published 2026-08-28 Alibaba study compares assistive and autonomous agent architectures and finds fundamentally different breakdown patterns. - **How Automation Reshapes the Queue That Reaches Human Workers** — [Source]() — Published 2026-08-28 Operations research formalizes how automated routing decisions change the composition of tasks humans actually handle. - **Half of Companies That Cut Service Staff for AI Will Rehire by 2027** — [Source]() — Published 2026-08-28 Industry data shows most organizations expanding human agent roles rather than eliminating them after automation. ### Past Articles - [What Aman Gupta's Agents Don't Decide]() — By [Sevda Polat]() · Published 2026-08-20 · Format: Builder Profile Aman Gupta's team at Nubank built an agent architecture for over a hundred million customers largely by subtraction: deterministic sequences collapsed into single tools, toolsets held to five or fifteen options, routines translated from human procedure. The model decides only what couldn't be engineered away in advance. That's a precise account of where the model's authority ends. What the architecture doesn't yet specify is what a person receives when a live case comes back to them, because the scratchpad that keeps the agent moving isn't the artifact someone needs to pick up where it stopped. - [The Intervention Horizon as a Design Parameter]() — By [Sable Whitford]() · Published 2026-08-20 · Format: Lead Analysis Agent architectures put a human approval step in the design and call it a safety measure. Approval only works if the person arrives before the outcome is locked in, with enough context to understand what they're approving and enough time to choose otherwise. Past that point, it records consent and offers no alternative. The last moment a human can still change what happens next deserves the same specification you'd write for latency, capacity, or failure behavior. Most systems never define it. - [Who Pays for the Record]() — By [Sevda Polat]() · Published 2026-08-12 · Format: Evidence Economics When an AI agent canceled a gym booking in Australia earlier this year, the only record of what happened was screenshots taken by the person who deployed the agent. The person whose reservation vanished had nothing. Records cost money to produce — instrumentation, storage, retrieval, access controls — and in agent systems that cost lands by default on whoever eventually needs one, which is rarely the party equipped to pay it. Regulated industries settled this question decades ago. Agent systems haven't, and the economics of the gap are entirely predictable. - [Melissa Pan's Three Evidence Problems]() — By [Sevda Polat]() · Published 2026-08-12 · Format: Builder Profile A database recovery log carries its own interpretation. A multi-agent failure trace does not: in one recent study, six experts had to invent the failure categories before 150 traces could be classified at all. And production agent deployments generate evidence that researchers can reach only by interviewing the people who work inside. ## Other sections in Issue 44 - [Daily Brief]() - [Market Pulse]() - [Foundations]() - [Echoes]() - [Vision]() ## More - [Publication guide]() - [Issue archive]()