Glossário de software e IA para executivos
693 termos de software e inteligência artificial, explicados para quem aprova, contrata e responde pelo resultado. Cada verbete diz o que o termo significa e o que conferir com a equipe. O glossário acompanha o curso Software e IA para executivos, e cada termo leva à aula que o explica.
Busca
693 termos. A busca aceita português ou o nome técnico, com ou sem acento.
Por letra
Sete categorias
Na ordem em que um produto digital se constrói e se governa.
Como o software é feito (94)
Papéis, stack, repositório, revisão, testes e cadência de entrega.
O que o cliente vê (71)
Telas, componentes, velocidade percebida e acessibilidade.
Regras e integrações (70)
O servidor, os contratos entre sistemas e o que acontece quando uma integração falha.
Dados (57)
Tabelas, consulta, índice, migração, cópia de segurança e isolamento entre clientes.
Segurança e acesso (64)
Quem entra, o que pode fazer, onde ficam os segredos e de onde vêm as dependências.
Publicação e operação (170)
Levar ao ar, voltar atrás, medir, pagar a conta e processar o que roda em segundo plano.
IA e agentes (167)
Modelos, fornecedores, contexto, ferramentas, avaliação, limites e custo por tarefa.
Outros glossários do portal
Todos os 693 termos
a
- ABAC (Attribute-Based Access Control)
- Access Token
- Action Policy
- Active CPU pricing
- Active LTS e Maintenance LTS
- Active-Active
- Active-Passive
- Activity
- Adapter Pattern
- Agent
- Agent debt (dívida de agente)
- Agent Sandbox
- Agentic AI Foundation
- Agentic Coding
- AGENTS.md
- AI Credits
- AI Safety Review
- Alert
- Analyze
- Angular
- Animation libraries
- Anonymization
- Answer Relevance
- Antigravity
- API
- API Gateway
- API Key
- API Versioning
- API-first
- APM
- Aposentadoria de modelo
- App Router
- AppArmor
- Approval
- Approval Workflow
- Architecture
- ARIA
- Armed
- Artifact
- At-Least-Once Delivery
- At-Most-Once Delivery
- Audit Log
- Auth
- Authentication
- Authentication (AuthN)
- Authorization
- Authorization (AuthZ)
- Auto mode (permissão por classificador)
- Autonomia Graduada
- Autonomous Agent
- Avro
- AWS
- Azure
b
- Backend
- Background Job
- Backoff
- Backpressure
- Backup Point-in-Time (PITR)
- Backward Compatibility
- Backward Compatible
- Batching
- Benchmark informado pelo fornecedor
- BFF
- Bias Test
- Blast Radius
- Blue-Green Deploy
- Body
- Bolt.new
- Bounded Context
- Branch
- Breaking Change
- Bucket
- Build
- Bulkhead
- Bun
- Bundle
- Bundle analysis
- Bundler
- Burst
- BYOK (Bring Your Own Key)
c
- Cache Components
- Cache-Control
- Canary Release
- Capabilities
- Causal Consistency
- CD
- CD (Continuous Delivery / Continuous Deployment)
- CDC
- CDC (Change Data Capture)
- CDC gerenciado
- CDN
- Chain-of-Thought/CoT
- Chain-of-Verification
- Changelog
- Chaos Engineering
- Check Constraint
- Cherry-pick
- Chunk Overlap
- Chunk Size
- CI
- CI (Continuous Integration)
- Circuit Breaker
- Claude Code
- Cláusula de mudança de controle
- Clean Architecture
- Client Component
- Clone
- CLS
- Cluster Autoscaler
- Cobrança por uso (ferramenta de IA)
- Code Review
- Code splitting
- Codebase
- Codex
- Coding Agent
- Coding Agent Index
- Cold Standby
- Cold Start
- Column
- Commit
- Commit Message
- Comms Lead
- Compaction
- Compensation
- Compilador nativo de TypeScript
- Completed
- Component
- Composite Key
- Conditional Request
- Confidence Score
- Connection Pool
- Consent Management
- Constraint
- Consumer Group
- Container
- Container na edge
- Context
- Context Engineering
- Context Poisoning
- Context Precision
- Context Window/Window
- Contract Testing
- Controller
- Correlation ID
- CORS
- CORS (Cross-Origin Resource Sharing)
- CPU Bound
- CQRS
- Credencial de curta duração
- Critic Agent
- Cron Job
- CSR
- CSRF
- CSRF (Cross-Site Request Forgery)
- CSS
- Cursor
- Cursor Pagination
- Curva J da adoção de IA
- Custo por tarefa
- Cycle Time
d
- Dashboard
- DAST
- DAST (Dynamic Application Security Testing)
- Data Catalog
- Data Contract
- Data Governance
- Data Lake
- Data Lineage
- Data Masking
- Data Minimization
- Data Retention
- Data Warehouse
- Database (DB)
- Dead Letter Queue/DLQ
- Definition of Done (DoD)
- Deno
- Denormalization
- Dependency Injection
- Dependency Pinning
- Deploy
- Deserialization
- Design System
- Deterministic Output
- DEV
- Dev Environment
- Diff
- Differential Privacy
- Dispatched
- Distributed Tracing
- Dívida técnica
- Docker
- Domain Model
- Downtime
- DPIA (Data Protection Impact Assessment)
- DR
- DTO
e
- E2E Test (End-to-End)
- Edge Function
- Edge Functions com state
- Edge Runtime
- Egress Control
- ELT
- Embedding
- Encryption
- Endpoint
- Engine
- Environment
- Environment Variables
- Episodic Memory
- Error Budget
- Error Tracking
- ETag
- ETL
- EU AI Act
- Eval
- Eval Harness
- Event
- Event Sourcing
- Event-Driven Architecture
- Eventual Consistency
- Exactly-Once Delivery
- Excessive Agency (agência excessiva)
- Execução durável
- Execution Plan
- Executor Agent
- Extensão gêmea maliciosa
i
- I/O Bound
- IaC
- IAM
- Idade mínima de versão (min-release-age)
- Idempotência em integração com IA
- Idempotency
- Idempotency Key
- Image Optimization
- Inbox Pattern
- Incident
- Incident Commander
- Index
- Indirect Prompt Injection
- Inference
- INP
- Instant Navigations e Partial Prefetching
- Integration Test
- Internationalization (i18n)
- Inversion of Control
- Islands Architecture
- ISR
l
- Lakehouse
- Latency
- Launch Checklist
- Lazy loading
- LCP
- Lead Time
- Leaky Bucket
- Least Privilege
- Leitura de cache (cache read)
- Letramento técnico
- LGPD Basis
- Lighthouse
- Lint
- Live
- Live Check
- Liveness Check
- LLM
- LLM Judge
- Local Model
- Localhost
- Localization (l10n)
- Lockfile
- Log
- Logging
- Long Polling
- Long-Term Memory
- Loose Coupling
- Lovable
m
- Main / Master
- Marshalling
- Materialized View
- MCP
- MCP Apps
- MCP Client
- MCP Extensions
- MCP Prompt
- MCP Resource
- MCP Server
- MCP tool poisoning
- Memory Conflict Resolution
- Memory Leak
- Memory Reflection
- Memory Store
- Merge
- Merge Conflict
- Merge queue (fila de merge)
- Metric
- Metrics
- MFA (Multi-Factor Authentication)
- Micro-frontend
- Microservices
- Middleware
- Migration
- Mock Server
- Model
- Model Card
- Model Cascade
- Model Drift
- Model Routing
- Modelo em Preview
- Modular Monolith
- Monorepo
- Motion design tokens
- MPA
- MTBF
- mTLS
- MTTR
- Multi-Region
- Multipart Upload
- MVP (Minimum Viable Product)
p
- P95/P99
- Pacote Alucinado
- Pagination
- Params
- Partition
- Pass Rate
- Passkey
- Path Params
- Payload
- Pending
- Permission Scope
- Persona Prompting
- PII (Personally Identifiable Information)
- Pipeline
- Placar independente
- Planner Agent
- Pluralization
- Polling
- Polyrepo
- PostgreSQL/Postgres
- Postmortem
- PR (Pull Request)
- PR per Hour
- Preload / Prefetch
- Preview
- Primary Key (PK)
- Prisma
- Privacy by Design
- Procedural Memory
- Production / Prod
- Production Hardening
- Projection
- Prompt
- Prompt Cache
- Prompt Drift
- Prompt Injection
- Protobuf
- Protocol Buffers
- Prototype
- Provenance
- Pseudonymization
- Pub/Sub
- Pull
- Push
- PWA
r
- Radix UI
- RAG
- Rate Limit
- Rate Limiting
- RBAC (Role-Based Access Control)
- Re-Ranker
- React
- ReAct
- React Query / TanStack Query
- Read Replica
- Read-Your-Writes Consistency
- Readiness Check
- Realtime
- Rebase
- Red Teaming
- Reduced Motion
- Reflexion
- Refresh Token
- Refusal Policy
- Regression Eval
- Regression Test
- Release
- REPACK CONCURRENTLY
- Replay
- Repo (Repository)
- Repository Pattern
- Request
- Request Changes
- Reserved Instance
- Responsabilidade objetiva por software
- Response
- Responsive Design
- REST
- REST maturity (modelo de Richardson)
- Retry
- Retry Logic
- Reversibilidade
- Review Comment
- Revisão assistida por IA
- Rework rate (taxa de retrabalho)
- RLS (Row Level Security)
- robots.txt
- Rollback
- Rolldown
- Root Cause
- Route
- Row
- RPO
- RTL support
- RTO
- Rug pull de servidor MCP
- Ruleset e branch protection
- RUM
- Runbook
- Runner
- Running
- Runtime
s
- Saga Compensation Order
- Saga Pattern
- Salt
- Sampling
- Sandbox API
- SAST
- SAST (Static Application Security Testing)
- Saturação de Benchmark
- Savings Plan
- SBOM
- SBOM (Software Bill of Materials)
- SCA
- Scale to Zero
- Scheduler
- Schema
- Schema Evolution
- Schema Registry
- Schema.org
- SDK
- Seccomp
- Secret
- Secret Management
- Secret Manager
- Secret Rotation
- Secrets Rotation
- Security Gate
- Security Patch
- Self-Consistency
- Self-Refine
- Semantic Memory
- Semantic Search
- Serialization
- Server
- Server Actions
- Server Component
- Server-Sent Events
- Serverless
- Service
- Service Boundary
- Service Mesh
- Service Worker
- Servidor MCP malicioso
- Session
- SEV
- shadcn/ui
- Short-Term Memory
- Signal
- Sitemap XML
- Skill (Agent Skill)
- SLA
- SLI
- SLO
- Slopsquatting
- Small Model Routing
- SOAP
- Sorting
- SPA
- Span
- Sparse Fieldsets
- Spec-Driven Development
- Spot Instance
- SQL
- SQL Injection
- Squash
- SRP
- SSE Transport
- SSG
- SSO (Single Sign-On)
- SSR
- Stack
- Staging
- Stale-While-Revalidate
- State management
- Static Analysis
- Status Page
- Stdio Transport
- Storage
- Stored Procedure / Function
- Strangler Fig Pattern
- Stream
- Stream Processing
- Streaming
- Streaming Responses
- Streaming SSR
- Strong Consistency
- Structured Output
- Subagente
- Supabase
- Supervisor Agent
- Supply Chain Attack
- Suspense
- Swagger
- SWE-bench Pro
- Synthetic Data
- System Card
- System Design
- System Prompt
t
- Table
- Tag
- Tailwind CSS
- Tailwind CSS v4
- Tech Stack
- Temperature
- Terminal-Bench
- Test Coverage
- Throttling
- Throughput
- Throughput de API
- Tier
- Timeout
- Token
- Token Bucket
- Token Budget
- Token Expiration
- Tombstone
- Tool Budget
- Tool Calling
- Tool Misuse
- Tool Permission
- Toxicity Score
- TPS
- Trace
- Trace Context
- Tracing
- Transaction
- Transactional Messaging
- Tree of Thoughts
- Tree-shaking
- Trigger
- Trigger de Banco
- tRPC
- Trunk-based development
- TTFT
- Two-Phase Commit/2PC
- Type Coercion
- TypeScript