Guides for teams putting AI to work

Practical answers to the questions that come up when a company funds, builds and runs AI, from choosing the first workflow and rolling out coding agents to testing AI systems, retrieval, model costs and platform security.

Planning and funding AI work

Choosing where AI goes first, what a pilot must prove, what a workflow really costs, who owns the result and when to change course.

PostTopic
AI business caseHow to work out whether AI pays off for one process: measure today's cost per case, count every cost of the AI system including the people who check it, and agree in advance when to stop. With a worksheet for your own figures.Business caseIntroducing AI in your companyWhere to start with AI, why so many pilots stall, and how one process gets from the first idea to a running system, with the people who need to be involved along the way.AI strategyAI readiness checkHow to test AI readiness on one real process by following a sample case through your systems. With a checklist and a fair look at what the free chamber checks cover.AI strategyCopilot or your own AI assistantCopilot covers work that lives in Microsoft 365. Your own AI assistant pays off when answers depend on an ERP, a DMS or data that must stay in-house. Either way, permissions come first.ProcurementChatGPT at work and data protectionFour ways to use ChatGPT-style AI at work, from private accounts to your own servers, and what leaves the company with each. Pick by the most sensitive data people need.SecurityShadow AI: ban or better optionShadow AI grows where companies encourage AI without providing it, and bans push it onto private phones. An approved platform that enforces your AI policy works better.GovernanceAI Act checklist for deployersWhat the EU AI Act asks of companies that use AI assistants and agents after the Digital Omnibus, with a technical checklist from inventory to supplier documentation.GovernanceAI agency, consultancy or own teamHow an AI agency, a consultancy, a freelancer and your own team differ, what to ask a partner for, and how a build with handover lets your own team take over later.ProcurementChange management for AIHow to introduce AI without losing your staff: an honest answer on jobs, the works council from the start, shadow mode with human approval and training on real work.RolloutAI pilot success criteriaWhat an AI pilot has to show before the next stage is funded, and how to write acceptance criteria before anyone chooses a tool.Business caseChoosing the first AI workflowHow to choose the first process to automate with AI by scoring every team's proposal on volume, variation, system access, review effort, cost of a mistake and known outcomes.AI strategyAI build or buyHow to decide between buying an AI product and building around the systems you already run, and what your team owns and maintains in each case.ProcurementAI project data readinessThe data owners, system access, representative cases and expert time an AI project needs before it starts, with a start checklist and what can wait.DeliveryFull cost of an AI workflowHow to compare AI options on the cost of a handled case, including review, escalations, rework and upkeep, with a worked example and its assumptions.Business caseHuman review budgetHow to budget reviewer hours before automating a process: review modes, a capacity sheet with peak days, faster reviews and when the review share may fall.OperationsAI pilot to productionWhat has to be in place before a successful AI pilot goes live: regression tests, versioning and rollback, monitoring, an owner, a handover and a staged rollout.DeliveryAI system ownershipWho owns the prompts, models, data, test set and failures of an AI system once the pilot team leaves, who approves which change and who can switch it off.OperationsAI vendor exit planWhat to keep in your own hands so that changing an AI vendor does not mean starting again: tests, prompts, rules, data, a dependency inventory and contract terms.ProcurementWhen to stop an AI projectHow to decide from measured quality and cost per case whether an AI project should continue, narrow its scope, switch to a simpler design or stop.AI strategyAI SDLCHow coding agents change each phase of the software development lifecycle, who owns each handoff, and how to measure and govern the change.ProcessAutonomous software factoryWhat it takes for coding agents to carry scoped work from issue to merge, which work fits, and where people stay in control.StrategyChoosing an AI consultancyWhat to ask a coding agent or AI consultancy before you hire it, and which evidence to request for each answer.Procurement

Rolling out coding agents

Pilots, repository setup, review, CI and team instructions for engineering teams that hand more of their work to coding agents.

PostTopic
Coding agents through an LLM gatewayWhat Claude Code, GitHub Copilot and Cursor each support when you put your own LLM gateway in front of them, as of November 2026, and what to log, budget and test before rollout.ToolingAgentic engineeringWhat agentic engineering means for an engineering team: how it differs from vibe coding, where the work moves, what the research says about results and which foundations decide whether it pays off.Delivery practiceCoding agent pilot designHow to compare coding agents on the same work, with random task assignment, a replay of merged changes and a decision rule agreed in advance.RolloutRepository readiness for coding agentsWhat a repository needs before coding agents can work in it reliably, and how to turn the setup instructions into a check that runs in CI.HarnessBranch protection for agent pull requestsThe GitHub branch protection and ruleset settings that should apply when coding agents open pull requests, and how to check that they actually hold.GovernanceAgent changes failing in CIHow to narrow down why a coding agent's change passed in its own environment and failed in CI, and which harness fixes stop it from recurring.HarnessRolling out Claude Code in a companyThe decisions an engineering lead makes before and during a Claude Code rollout: accounts and provider, managed settings and permissions, what goes into the repository, a pilot with one team, training and the works council.Claude CodeClaude Code team instructionsHow a team assigns owners to CLAUDE.md files and rules, reviews them, and retires entries that no longer change what Claude Code does.Claude CodeCursor rules in a monorepoHow to keep Cursor rules relevant to each package in a monorepo, with path-scoped rules, nested AGENTS.md files and an owner for every rule.CursorGitHub Copilot pilot metricsWhat GitHub's Copilot usage metrics measure, where they stop, and how to combine them with your own review and rework data to decide whether to continue.GitHub CopilotHanding tasks to cloud coding agentsWhat a task should contain before it goes to a cloud coding agent, with a reusable brief, three readiness questions and rules for handing work back.WorkflowCoding agent context budgetsHow to reduce the context a coding agent carries without dropping constraints that matter, and how to check on your own tasks that quality held.CostTesting coding agent upgradesHow to check whether a new model or agent version is better on your repositories before it becomes the team default, with pinned versions and a paired comparison.EvaluationSpec-driven developmentHow to write specs coding agents can follow, take one change from spec to merge, and choose between Spec Kit, OpenSpec, BMad, GSD and Kiro.Delivery practiceAI code reviewHow to set up AI code review and keep a person accountable for every merge when coding agents write more of your pull requests.Qualitypstack workflowsWhat the pstack plugin for Cursor and Claude Code contains, how its playbooks and model roles work, and how a team adopts it under its own controls.ToolingCoding agents compared for teamsClaude Code, Cursor, GitHub Copilot and OpenAI Codex side by side: plans, usage billing, models, instruction files, data handling and when each one fits.ToolingWhat a coding agent rollout costsSeat prices for 50 engineers on Claude Code, Cursor, GitHub Copilot and Codex, why usage and review time often cost more, and how to budget per merged change.CostsCoding agent cost calculatorSeats, usage beyond the seat and review time for your own team at the vendors' list prices, with the cost per merged change worked out for you.Costs

Process agents in operation

Testing, approvals, escalation, monitoring and recovery for agents that handle business cases and act in your systems.

PostTopic
Agent test sets from past casesHow to pick past business cases for a process agent's test set and keep a held-out part that nobody tunes against.AI agentsOutcome versus trace evaluationWhy an agent run can finish without errors and still handle the case wrongly, and how to check business outcomes separately from traces.AI agentsHuman approval in agent workflowsWhich agent actions on business records need a person's approval, what the approver must see and how to keep approvals from stalling every case.AI agentsRecovering from agent tool failuresWhat a process agent should do when an API times out or a tool returns partial data, with typed failures, bounded retries and a clean handoff.AI agentsIdempotency for agent actionsWhere duplicate payments and records come from in agent systems, how to key each business action and why the receiving system must enforce it.AI agentsMonitoring a live process agentThe signals that show a live process agent getting more expensive or handling cases worse, measured per case and tied to objectives you agree in advance.AI agentsDesigning agent escalationsWhen a process agent should hand a case to a person, what the handoff must contain and how to run the queue that receives it.AI agentsShadow mode for process agentsHow to run a process agent on live cases without letting it change anything, compare its proposals with people's decisions and decide when it may act.AI agentsVersioning a process agentWhat needs a version when a process agent changes, how to stamp it on every case and how to explain a result months later without rerunning it.AI agentsReviewing agent exceptionsHow to review a production agent's exceptions with a fixed time budget, triage them to an owner and turn the right ones into regression cases.AI agents

Answers from your documents

Preparing sources, keeping permissions intact, checking citations and keeping retrieval assistants accurate as documents change.

PostTopic
RAG source readinessHow to decide which documents a retrieval assistant should use, who owns each source and how to find missing answers before anything is indexed.RAGRAG document permissionsHow access rules travel from source systems into a retrieval index, where they get lost, and which negative tests show that a user cannot retrieve what they may not open.RAGRAG citation qualityHow to test whether the passages an assistant cites actually support each claim it makes, with a claim-level check, a worked example and acceptance criteria.RAGRetrieval or generation?A diagnostic procedure for finding out whether a wrong RAG answer started in ingestion, search, reranking or the model, using labelled evidence and per-question traces.RAGStale documents in RAGHow to keep a retrieval assistant current when documents are edited, renamed, restricted or deleted, with a sync design and canary tests that prove old content is gone.RAGChunking business documentsHow to split contracts, policies and long procedures so that an answer stays with its exceptions and definitions, and how to choose a strategy by experiment on your own questions.RAGRAG or fine-tuning?Decision criteria for choosing retrieval, fine-tuning or both: whether the task needs changing knowledge or consistent behaviour, with a worked table and an order of work.RAGTables and spreadsheets in RAGWhen an assistant should retrieve table rows as text and when it should run a query instead, with routing rules, row formatting for search and safeguards for generated SQL.RAGMultilingual RAG testingHow to test whether an assistant finds the same evidence when people ask in English, German or Polish, with a parallel test set, a language matrix and terminology fixes.RAGRAG production handoverWhat a RAG supplier should hand over so your team can run and change the system: source connectors, configuration, evaluation assets, runbooks, owners and a rehearsal.RAG

Models, inference and cost

When to fine-tune or use a smaller model, how to test it against the current one, and where latency and model spend come from.

PostTopic
Fine-tuning data qualityFour checks that show whether your examples are ready for fine-tuning: label errors, coverage of real traffic, provenance and leakage into the test set.ModelsFine-tuned model evaluationHow to show that a fine-tuned model should replace the current one: acceptance criteria written in advance, a fair held-out comparison, and cost and latency in the same table.ModelsSmall model task selectionHow to find the high-volume tasks a small language model can handle reliably: a screening checklist, a test before any training and a cost comparison at your own volume.ModelsModel distillation trade-offsWhen distilling a large model into a smaller one is worth the training work: what the teacher costs, which method needs what, the quality you can lose and how to test for it.ModelsLLM latency budgetHow to find out why an AI feature is slow: split each request into queueing, retrieval, model and tool spans, then fix the largest one you can change and measure the tail under load.PerformanceLLM routing quality controlsHow to send easy requests to cheaper models without hiding quality losses: test the router like a classifier, keep fallbacks separate and log which model actually answered.CostsPrompt caching savingsWhy prompt caching saves less than expected and how to measure it: eligibility rules, the usual causes of cache misses, per-request usage fields and reconciliation with the bill.CostsGPU serving capacity planningHow to size GPU capacity for a model you serve yourself: what limits a replica, how to benchmark it at your latency target, how many replicas peak traffic needs and how to avoid paying for idle hours.InfrastructureQuantisation quality checksWhich tests to rerun before a quantised model replaces the original: task quality, structured output, long inputs, behaviour checks and serving performance, with cost compared per GPU.ModelsLLM cost allocationHow to explain an LLM bill by team, task and accepted outcome: tag requests at the gateway, price them as you are invoiced, allocate shared costs by a stated rule and reconcile with finance every month.Costs

Platform, integration and security

Connecting AI to enterprise systems through MCP and identity, containing prompt injection, handling credentials and running AI services reliably.

PostTopic
On-premise LLM for companiesWhat it takes to run an LLM on your own servers so that data stays in the company, and when a model in your own cloud account is enough, with much less to operate.InfrastructureWhat an LLM gateway doesWhat an LLM gateway (also called an AI gateway or LLM proxy) takes off your applications and what it leaves to the rest of the platform, with the main options compared as of 7 November 2026.InfrastructureLiteLLM in productionHow to keep running a self-hosted LiteLLM gateway after the March and October 2026 release compromises and CVE-2026-42271, from checking whether you were hit to the settings and checklist we use before go-live.SecurityEU data residency on Azure OpenAI and BedrockWhich Azure OpenAI deployment types and Amazon Bedrock inference profiles keep model requests in the EU, who can still read prompts, and how one gateway holds every team to EU endpoints. An EU region does not take the provider out of CLOUD Act scope.SecurityWorks council and AI logsWhich fields in the logs of an AI assistant, an LLM gateway or a coding agent make monitoring possible under German co-determination law, and how to set up logging so a works agreement holds.OperationsMCP server permissionsHow to connect an assistant to internal systems without giving every MCP tool broad access: scopes, checks inside each tool and separate downstream identities.SecurityMCP tool designWhy an MCP server should usually offer a few tools shaped around business actions instead of mirroring every API endpoint, with a worked example and a tool contract.IntegrationMCP and enterprise identityWhose identity an MCP tool should use when it calls a business system: user delegation, token exchange or a service identity, with the token boundaries to check.SecurityMCP server test planA test plan for an MCP server before users connect: contract tests, negative authorisation tests, failure tests and state-change tests, with checks that run in CI.QualityPrompt injection boundariesHow to limit the damage when an agent follows instructions hidden in a document, email or tool result: map untrusted inputs to actions, cut the paths and test the boundaries.SecuritySelf-hosted AI or managed APIHow to decide whether an AI workload should run on your own infrastructure or stay on a managed model endpoint, using access requirements, operating capacity and measured costs.InfrastructureAI trace data retentionWhat to log to debug agents without keeping every sensitive prompt: three tiers of trace data, content capture off by default, the limits of redaction and retention set by your policy owners.OperationsRunbooks for AI featuresWhat on-call engineers should do when an AI feature is slow, unavailable, expensive, wrong or acting outside its limits: symptom classes, runbook entries, safe switches and owners.ReliabilityModel provider failoverHow to fail over to another model during an outage without silently changing quality, permissions or data terms: a fallback contract, switch triggers, equivalence tests and rollback.ReliabilityAgent secrets and credentialsWhere credentials used by agents and model integrations should live, how to give each agent its own identity, how to rotate without breaking running workflows and what to do after a leak.SecurityCoding agent securityPractical security and governance for coding agents: prompt injection, secrets, permissions, sandboxing, unattended runs and ownership.Security

Dates and sources

Models, tools and vendor terms change often, so each guide shows the date it was last checked and lists the sources behind it. What a guide says about a product applies to the versions and documentation it cites.

If something has changed since, tell us and we will update the guide.

Talk to an engineer

Bring one piece of work you would like to improve. The call takes 30 minutes and needs no access to your systems or your data.

Booking runs on Google Calendar.

Mickey Graf

Mickey Graf

Founder and CEO of Cloudsail

Rated 4.9 out of 5 on Clutch

Prefer to write first? Leave your email and a few lines. Mickey reads every message himself and replies by email with times for a call.

The button prepares an email in your own email program, for you to send. Privacy policy (draft)