Multilingual agent engineering

Agent architectures that preserve locale, language, meaning, and policy across prompts, memory, retrieval, tools, and actions. In practice, the work is bounded by a defined product or model decision, named audiences and locales, representative inputs, and acceptance criteria that can be reviewed.

An agent can answer in the requested language while silently losing locale in memory, querying the wrong-language corpus, mistranslating tool parameters, or executing an action under the wrong market rules. The useful starting point is the smallest representative flow that can expose the cause, impact, and ownership of the problem.

Model the agent loop: Map prompts, state, memory, retrieval, tools, modalities, actions, policies, and every point where locale can be dropped or inferred. Engineer explicit contracts: Define locale state, schemas, language-aware routing, evidence handling, tool safeguards, and recovery behavior. Evaluate trajectories: Test multi-step tasks across languages, code-switching, modality changes, tool failures, and consequential actions.

The most useful inputs are the product task and user journey, candidate models or system access, priority languages and communities, policies and risk thresholds, representative prompts, media, tools, and and expected outcomes. Admas can begin with a partial package, but missing context, rights, access, owners, or acceptance criteria will be made visible in the plan rather than treated as harmless assumptions.

Typical outputs include multilingual agent reference architecture, locale-state and tool-contract specification, representative multilingual trajectory suite, and release gates and failure-recovery guidance. Deliverables are adapted to the team that must use them, with decisions, evidence, limitations, owners, and next actions made explicit.

Quality is measured against the real task and risk. Relevant evidence can include task success by language and scenario, human-rated meaning and usefulness, safety and policy performance, retrieval and citation fidelity, tool-call correctness, and regressions and disparities hidden by aggregate scores. Sampling, severity rules, reviewers, adjudication, and pass or fail thresholds should be agreed before the result is used as a release decision.

Models and automated checks can generate candidates, expand test sets, cluster failures, and accelerate analysis. Qualified humans define what good means, identify culturally or linguistically plausible failures, adjudicate close cases, and own consequential release judgments. The right allocation depends on consequence, content stability, available references, language coverage, reversibility, and the cost of a plausible but wrong result.

The estimate changes with number of languages, modalities, systems, and scenarios, risk level, dataset creation needs, evaluator specialization, sampling and adjudication depth, and experiment and reporting cadence. Pricing should distinguish setup and discovery, repeatable units, specialist or engineering time, independent review, management, and external costs. A low unit price is not comparable if it excludes the QA cycle or shifts rework back to the buyer.