AI Engineer 100 · AI Coding001 / 100
AI Coding#001

GitHub Copilot

AI coding assistant named explicitly in the brief for accelerating software delivery and developer productivity.

Role lens

Use AI coding tools inside normal engineering controls: scope, tests, review and release.

Concept diagram
Repository + intent
GitHub Copilot
Faster safe delivery
AI coding assistant named explicitly in the brief for accelerating software delivery and developer productivity.
Back to role-fit map
AI Coding#002

Claude Code

Agentic coding environment for repo-aware analysis, implementation, testing, refactoring and engineering automation.

Role lens

Use AI coding tools inside normal engineering controls: scope, tests, review and release.

Concept diagram
Repository + intent
Claude Code
Faster safe delivery
Agentic coding environment for repo-aware analysis, implementation, testing, refactoring and engineering automation.
Back to role-fit map
LLMs + Prompting#003

LLMs

Large Language Models used as the reasoning and generation layer inside production AI applications.

Role lens

Treat prompts and model choice as testable production inputs, not one-off chat experiments.

Concept diagram
Context + instruction
LLMs
Reliable model output
Large Language Models used as the reasoning and generation layer inside production AI applications.
Back to role-fit map
LLMs + Prompting#004

Generative AI

Applications that generate, transform, reason over or act on content using foundation models.

Role lens

Treat prompts and model choice as testable production inputs, not one-off chat experiments.

Concept diagram
Context + instruction
Generative AI
Reliable model output
Applications that generate, transform, reason over or act on content using foundation models.
Back to role-fit map
LLMs + Prompting#005

OpenAI GPT

OpenAI foundation models used through APIs for reasoning, generation and application workflows.

Role lens

Treat prompts and model choice as testable production inputs, not one-off chat experiments.

Concept diagram
Context + instruction
OpenAI GPT
Reliable model output
OpenAI foundation models used through APIs for reasoning, generation and application workflows.
Back to role-fit map
LLMs + Prompting#006

Azure OpenAI

Enterprise Azure-hosted access to OpenAI models with Microsoft cloud governance and integration.

Role lens

Treat prompts and model choice as testable production inputs, not one-off chat experiments.

Concept diagram
Context + instruction
Azure OpenAI
Reliable model output
Enterprise Azure-hosted access to OpenAI models with Microsoft cloud governance and integration.
Back to role-fit map
LLMs + Prompting#007

Anthropic Claude

Anthropic model family used for reasoning, coding, agents and enterprise AI applications.

Role lens

Treat prompts and model choice as testable production inputs, not one-off chat experiments.

Concept diagram
Context + instruction
Anthropic Claude
Reliable model output
Anthropic model family used for reasoning, coding, agents and enterprise AI applications.
Back to role-fit map
LLMs + Prompting#008

Claude APIs

Programmatic integration of Claude models into services, applications and agent workflows.

Role lens

Treat prompts and model choice as testable production inputs, not one-off chat experiments.

Concept diagram
Context + instruction
Claude APIs
Reliable model output
Programmatic integration of Claude models into services, applications and agent workflows.
Back to role-fit map
LLMs + Prompting#009

Prompt Engineering

Designing instructions, context and output constraints so models behave reliably for a defined task.

Role lens

Treat prompts and model choice as testable production inputs, not one-off chat experiments.

Concept diagram
Context + instruction
Prompt Engineering
Reliable model output
Designing instructions, context and output constraints so models behave reliably for a defined task.
Back to role-fit map
LLMs + Prompting#010

Prompt Optimisation

Iteratively improving prompts against measurable quality, cost, latency and failure criteria.

Role lens

Treat prompts and model choice as testable production inputs, not one-off chat experiments.

Concept diagram
Context + instruction
Prompt Optimisation
Reliable model output
Iteratively improving prompts against measurable quality, cost, latency and failure criteria.
Back to role-fit map
Agents + SDLC#011

AI Agents

LLM-powered components that reason, select tools and execute bounded multi-step tasks.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
AI Agents
Bounded action
LLM-powered components that reason, select tools and execute bounded multi-step tasks.
Back to role-fit map
Agents + SDLC#012

Agentic AI

Systems where models can plan and take controlled actions rather than only return text.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
Agentic AI
Bounded action
Systems where models can plan and take controlled actions rather than only return text.
Back to role-fit map
Agents + SDLC#013

Copilots

AI assistants embedded into human workflows to accelerate decisions, coding or operational tasks.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
Copilots
Bounded action
AI assistants embedded into human workflows to accelerate decisions, coding or operational tasks.
Back to role-fit map
Agents + SDLC#014

Workflow Automation

Connecting AI reasoning to deterministic business and engineering actions through tools and APIs.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
Workflow Automation
Bounded action
Connecting AI reasoning to deterministic business and engineering actions through tools and APIs.
Back to role-fit map
Agents + SDLC#015

LangChain

Framework for composing LLM calls, retrieval, tools and application workflows.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
LangChain
Bounded action
Framework for composing LLM calls, retrieval, tools and application workflows.
Back to role-fit map
Agents + SDLC#016

LangGraph

Graph-based orchestration framework for stateful, multi-step and agentic LLM workflows.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
LangGraph
Bounded action
Graph-based orchestration framework for stateful, multi-step and agentic LLM workflows.
Back to role-fit map
Agents + SDLC#017

CrewAI

Multi-agent framework for assigning roles and coordinating collaborative AI tasks.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
CrewAI
Bounded action
Multi-agent framework for assigning roles and coordinating collaborative AI tasks.
Back to role-fit map
Agents + SDLC#018

AutoGen

Microsoft-originated framework for conversational and multi-agent orchestration patterns.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
AutoGen
Bounded action
Microsoft-originated framework for conversational and multi-agent orchestration patterns.
Back to role-fit map
Agents + SDLC#019

Semantic Kernel

Microsoft SDK for integrating models, plugins, memory and orchestration into applications.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
Semantic Kernel
Bounded action
Microsoft SDK for integrating models, plugins, memory and orchestration into applications.
Back to role-fit map
Agents + SDLC#020

Code Generation

Using AI to produce or modify production code within normal engineering quality controls.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
Code Generation
Bounded action
Using AI to produce or modify production code within normal engineering quality controls.
Back to role-fit map
Agents + SDLC#021

AI-assisted SDLC

Applying AI across analysis, coding, tests, review, documentation, release and maintenance.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
AI-assisted SDLC
Bounded action
Applying AI across analysis, coding, tests, review, documentation, release and maintenance.
Back to role-fit map
Agents + SDLC#022

Developer Productivity

Reducing engineering cycle time while protecting quality, maintainability and operational safety.

Role lens

Agents should plan, use tools and automate work while remaining observable, testable and controlled.

Concept diagram
Goal + workflow
Developer Productivity
Bounded action
Reducing engineering cycle time while protecting quality, maintainability and operational safety.
Back to role-fit map
Software Engineering#023

Python

Primary language for AI services, orchestration, data processing and backend application development.

Role lens

The AI layer still depends on strong APIs, languages, architecture and scalable service boundaries.

Concept diagram
Typed contract
Python
Production service
Primary language for AI services, orchestration, data processing and backend application development.
Back to role-fit map
Software Engineering#024

FastAPI

Python web framework commonly used for typed, high-performance AI and microservice APIs.

Role lens

The AI layer still depends on strong APIs, languages, architecture and scalable service boundaries.

Concept diagram
Typed contract
FastAPI
Production service
Python web framework commonly used for typed, high-performance AI and microservice APIs.
Back to role-fit map
Software Engineering#025

JavaScript

Core web and application language named as an acceptable software-engineering skill in the brief.

Role lens

The AI layer still depends on strong APIs, languages, architecture and scalable service boundaries.

Concept diagram
Typed contract
JavaScript
Production service
Core web and application language named as an acceptable software-engineering skill in the brief.
Back to role-fit map
Software Engineering#026

TypeScript

Typed JavaScript used for maintainable frontend, Node.js and full-stack production systems.

Role lens

The AI layer still depends on strong APIs, languages, architecture and scalable service boundaries.

Concept diagram
Typed contract
TypeScript
Production service
Typed JavaScript used for maintainable frontend, Node.js and full-stack production systems.
Back to role-fit map
Software Engineering#027

C#

Enterprise application language explicitly accepted by the role alongside Python, JavaScript/TypeScript and Java.

Role lens

The AI layer still depends on strong APIs, languages, architecture and scalable service boundaries.

Concept diagram
Typed contract
C#
Production service
Enterprise application language explicitly accepted by the role alongside Python, JavaScript/TypeScript and Java.
Back to role-fit map
Software Engineering#028

Java

Enterprise backend language explicitly accepted by the role for production application engineering.

Role lens

The AI layer still depends on strong APIs, languages, architecture and scalable service boundaries.

Concept diagram
Typed contract
Java
Production service
Enterprise backend language explicitly accepted by the role for production application engineering.
Back to role-fit map
Software Engineering#029

REST APIs

HTTP interfaces used to expose AI capabilities and connect models with enterprise systems.

Role lens

The AI layer still depends on strong APIs, languages, architecture and scalable service boundaries.

Concept diagram
Typed contract
REST APIs
Production service
HTTP interfaces used to expose AI capabilities and connect models with enterprise systems.
Back to role-fit map
Software Engineering#030

Microservices

Independently deployable service boundaries used to integrate AI safely into larger platforms.

Role lens

The AI layer still depends on strong APIs, languages, architecture and scalable service boundaries.

Concept diagram
Typed contract
Microservices
Production service
Independently deployable service boundaries used to integrate AI safely into larger platforms.
Back to role-fit map
Software Engineering#031

Scalable Architecture

Designing applications so model, data, API and infrastructure layers can scale independently.

Role lens

The AI layer still depends on strong APIs, languages, architecture and scalable service boundaries.

Concept diagram
Typed contract
Scalable Architecture
Production service
Designing applications so model, data, API and infrastructure layers can scale independently.
Back to role-fit map
Software Engineering#032

Cloud-native

Building services around managed cloud infrastructure, automation, observability and elastic deployment.

Role lens

The AI layer still depends on strong APIs, languages, architecture and scalable service boundaries.

Concept diagram
Typed contract
Cloud-native
Production service
Building services around managed cloud infrastructure, automation, observability and elastic deployment.
Back to role-fit map
Software Engineering#033

Enterprise Integration

Connecting AI to real identity, APIs, data, workflows and systems of record rather than isolated demos.

Role lens

The AI layer still depends on strong APIs, languages, architecture and scalable service boundaries.

Concept diagram
Typed contract
Enterprise Integration
Production service
Connecting AI to real identity, APIs, data, workflows and systems of record rather than isolated demos.
Back to role-fit map
RAG + Knowledge#034

RAG

Retrieval-Augmented Generation: grounding model responses in retrieved enterprise knowledge.

Role lens

Retrieval quality, permissions and provenance determine whether enterprise AI can be trusted.

Concept diagram
Enterprise knowledge
RAG
Grounded answer
Retrieval-Augmented Generation: grounding model responses in retrieved enterprise knowledge.
Back to role-fit map
RAG + Knowledge#035

Embeddings

Vector representations of text or data used for semantic retrieval and similarity search.

Role lens

Retrieval quality, permissions and provenance determine whether enterprise AI can be trusted.

Concept diagram
Enterprise knowledge
Embeddings
Grounded answer
Vector representations of text or data used for semantic retrieval and similarity search.
Back to role-fit map
RAG + Knowledge#036

Semantic Search

Retrieving information by meaning rather than only exact keyword matching.

Role lens

Retrieval quality, permissions and provenance determine whether enterprise AI can be trusted.

Concept diagram
Enterprise knowledge
Semantic Search
Grounded answer
Retrieving information by meaning rather than only exact keyword matching.
Back to role-fit map
RAG + Knowledge#037

Vector Database

Storage and search layer optimised for embeddings and similarity-based retrieval.

Role lens

Retrieval quality, permissions and provenance determine whether enterprise AI can be trusted.

Concept diagram
Enterprise knowledge
Vector Database
Grounded answer
Storage and search layer optimised for embeddings and similarity-based retrieval.
Back to role-fit map
RAG + Knowledge#038

Pinecone

Managed vector database named in the job description for RAG and semantic retrieval.

Role lens

Retrieval quality, permissions and provenance determine whether enterprise AI can be trusted.

Concept diagram
Enterprise knowledge
Pinecone
Grounded answer
Managed vector database named in the job description for RAG and semantic retrieval.
Back to role-fit map
RAG + Knowledge#039

ChromaDB

Open-source embedding database named in the brief for vector search and RAG workflows.

Role lens

Retrieval quality, permissions and provenance determine whether enterprise AI can be trusted.

Concept diagram
Enterprise knowledge
ChromaDB
Grounded answer
Open-source embedding database named in the brief for vector search and RAG workflows.
Back to role-fit map
RAG + Knowledge#040

Weaviate

Vector database and semantic-search platform explicitly listed in the role requirements.

Role lens

Retrieval quality, permissions and provenance determine whether enterprise AI can be trusted.

Concept diagram
Enterprise knowledge
Weaviate
Grounded answer
Vector database and semantic-search platform explicitly listed in the role requirements.
Back to role-fit map
RAG + Knowledge#041

Azure AI Search

Azure search service supporting enterprise retrieval, hybrid search and vector search.

Role lens

Retrieval quality, permissions and provenance determine whether enterprise AI can be trusted.

Concept diagram
Enterprise knowledge
Azure AI Search
Grounded answer
Azure search service supporting enterprise retrieval, hybrid search and vector search.
Back to role-fit map
RAG + Knowledge#042

Knowledge Repositories

Authoritative document, database and content sources used to ground AI applications.

Role lens

Retrieval quality, permissions and provenance determine whether enterprise AI can be trusted.

Concept diagram
Enterprise knowledge
Knowledge Repositories
Grounded answer
Authoritative document, database and content sources used to ground AI applications.
Back to role-fit map
Evals + Governance#043

Model Evaluation

Testing model outputs against task-specific quality, correctness and reliability criteria.

Role lens

Quality and safety must be measured through evals, traces, monitoring, policy and human oversight.

Concept diagram
AI behaviour
Model Evaluation
Evidence + trust
Testing model outputs against task-specific quality, correctness and reliability criteria.
Back to role-fit map
Evals + Governance#044

LLM Evals

Repeatable automated or human evaluation sets used to detect regressions in AI behaviour.

Role lens

Quality and safety must be measured through evals, traces, monitoring, policy and human oversight.

Concept diagram
AI behaviour
LLM Evals
Evidence + trust
Repeatable automated or human evaluation sets used to detect regressions in AI behaviour.
Back to role-fit map
Evals + Governance#045

Observability

Tracing model calls, retrieval, tools, errors, latency and cost across production AI workflows.

Role lens

Quality and safety must be measured through evals, traces, monitoring, policy and human oversight.

Concept diagram
AI behaviour
Observability
Evidence + trust
Tracing model calls, retrieval, tools, errors, latency and cost across production AI workflows.
Back to role-fit map
Evals + Governance#046

Model Monitoring

Watching production AI quality, drift, failures, latency, usage and operational health over time.

Role lens

Quality and safety must be measured through evals, traces, monitoring, policy and human oversight.

Concept diagram
AI behaviour
Model Monitoring
Evidence + trust
Watching production AI quality, drift, failures, latency, usage and operational health over time.
Back to role-fit map
Evals + Governance#047

AI Governance

Policies and controls defining acceptable AI use, accountability, auditability and change management.

Role lens

Quality and safety must be measured through evals, traces, monitoring, policy and human oversight.

Concept diagram
AI behaviour
AI Governance
Evidence + trust
Policies and controls defining acceptable AI use, accountability, auditability and change management.
Back to role-fit map
Evals + Governance#048

Responsible AI

Designing AI systems around safety, transparency, privacy, fairness and accountable human oversight.

Role lens

Quality and safety must be measured through evals, traces, monitoring, policy and human oversight.

Concept diagram
AI behaviour
Responsible AI
Evidence + trust
Designing AI systems around safety, transparency, privacy, fairness and accountable human oversight.
Back to role-fit map
Evals + Governance#049

Security

Protecting identities, secrets, data, tools and model boundaries with explicit access controls.

Role lens

Quality and safety must be measured through evals, traces, monitoring, policy and human oversight.

Concept diagram
AI behaviour
Security
Evidence + trust
Protecting identities, secrets, data, tools and model boundaries with explicit access controls.
Back to role-fit map
Evals + Governance#050

Human-in-the-loop

Requiring human review or approval for higher-risk AI decisions and actions.

Role lens

Quality and safety must be measured through evals, traces, monitoring, policy and human oversight.

Concept diagram
AI behaviour
Human-in-the-loop
Evidence + trust
Requiring human review or approval for higher-risk AI decisions and actions.
Back to role-fit map
Cloud + Enterprise#051

AWS

Cloud platform listed as desirable for deploying and operating production AI applications.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
AWS
Operated platform
Cloud platform listed as desirable for deploying and operating production AI applications.
Back to role-fit map
Cloud + Enterprise#052

Azure

Microsoft cloud platform relevant to Azure OpenAI, enterprise identity, DevOps and AI services.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
Azure
Operated platform
Microsoft cloud platform relevant to Azure OpenAI, enterprise identity, DevOps and AI services.
Back to role-fit map
Cloud + Enterprise#053

GCP

Google Cloud Platform, included among the desirable cloud-platform skills in the brief.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
GCP
Operated platform
Google Cloud Platform, included among the desirable cloud-platform skills in the brief.
Back to role-fit map
Cloud + Enterprise#054

GitHub Actions

CI/CD automation for testing, validating and releasing software directly from GitHub workflows.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
GitHub Actions
Operated platform
CI/CD automation for testing, validating and releasing software directly from GitHub workflows.
Back to role-fit map
Cloud + Enterprise#055

Azure DevOps

Microsoft platform for repos, pipelines, work tracking and enterprise software delivery.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
Azure DevOps
Operated platform
Microsoft platform for repos, pipelines, work tracking and enterprise software delivery.
Back to role-fit map
Cloud + Enterprise#056

Jenkins

CI/CD automation server listed as a desirable engineering-delivery technology.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
Jenkins
Operated platform
CI/CD automation server listed as a desirable engineering-delivery technology.
Back to role-fit map
Cloud + Enterprise#057

Docker

Container technology for packaging AI services and their dependencies consistently across environments.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
Docker
Operated platform
Container technology for packaging AI services and their dependencies consistently across environments.
Back to role-fit map
Cloud + Enterprise#058

Kubernetes

Container orchestration platform for scheduling, scaling and operating distributed services.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
Kubernetes
Operated platform
Container orchestration platform for scheduling, scaling and operating distributed services.
Back to role-fit map
Cloud + Enterprise#059

Terraform

Infrastructure-as-code tooling for reproducible cloud environments and platform configuration.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
Terraform
Operated platform
Infrastructure-as-code tooling for reproducible cloud environments and platform configuration.
Back to role-fit map
Cloud + Enterprise#060

CI/CD

Automated build, test, security and deployment controls that move AI code safely into production.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
CI/CD
Operated platform
Automated build, test, security and deployment controls that move AI code safely into production.
Back to role-fit map
Cloud + Enterprise#061

APIs

The service contracts through which AI capabilities, enterprise data and tools are integrated.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
APIs
Operated platform
The service contracts through which AI capabilities, enterprise data and tools are integrated.
Back to role-fit map
Cloud + Enterprise#062

Mentoring

Helping developers adopt AI-assisted engineering patterns rather than creating dependence on one specialist.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
Mentoring
Operated platform
Helping developers adopt AI-assisted engineering patterns rather than creating dependence on one specialist.
Back to role-fit map
Cloud + Enterprise#063

Stakeholder Use Cases

Turning business needs into prioritised AI opportunities with measurable outcomes and delivery constraints.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
Stakeholder Use Cases
Operated platform
Turning business needs into prioritised AI opportunities with measurable outcomes and delivery constraints.
Back to role-fit map
Cloud + Enterprise#064

Enterprise AI

Production AI built within real organisational constraints: data, identity, governance, reliability and operations.

Role lens

Enterprise AI has to survive identity, cloud, CI/CD, containers, stakeholders and real operational constraints.

Concept diagram
Build artefact
Enterprise AI
Operated platform
Production AI built within real organisational constraints: data, identity, governance, reliability and operations.
Back to role-fit map
Agent Engineering#065

MCP

Model Context Protocol: a standard way to expose typed tools and context to coding agents and AI applications.

Role lens

Reusable tools, MCP, state and delegation turn an LLM into a controlled engineering agent.

Concept diagram
Agent state
MCP
Tool-backed action
Model Context Protocol: a standard way to expose typed tools and context to coding agents and AI applications.
Back to role-fit map
Agent Engineering#066

Tool Calling

Allowing a model or agent to select and invoke approved external functions, APIs or services.

Role lens

Reusable tools, MCP, state and delegation turn an LLM into a controlled engineering agent.

Concept diagram
Agent state
Tool Calling
Tool-backed action
Allowing a model or agent to select and invoke approved external functions, APIs or services.
Back to role-fit map
Agent Engineering#067

Function Calling

Structured model output that maps natural-language intent into deterministic function arguments.

Role lens

Reusable tools, MCP, state and delegation turn an LLM into a controlled engineering agent.

Concept diagram
Agent state
Function Calling
Tool-backed action
Structured model output that maps natural-language intent into deterministic function arguments.
Back to role-fit map
Agent Engineering#068

Multi-agent Systems

Architectures where specialised agents coordinate through explicit roles, hand-offs and shared state.

Role lens

Reusable tools, MCP, state and delegation turn an LLM into a controlled engineering agent.

Concept diagram
Agent state
Multi-agent Systems
Tool-backed action
Architectures where specialised agents coordinate through explicit roles, hand-offs and shared state.
Back to role-fit map
Agent Engineering#069

Router / Planner / Executor

Common agent pattern that separates task classification, decomposition and action execution.

Role lens

Reusable tools, MCP, state and delegation turn an LLM into a controlled engineering agent.

Concept diagram
Agent state
Router / Planner / Executor
Tool-backed action
Common agent pattern that separates task classification, decomposition and action execution.
Back to role-fit map
Agent Engineering#070

Memory + State

Durable conversation or workflow state that lets agents continue tasks safely across multiple steps.

Role lens

Reusable tools, MCP, state and delegation turn an LLM into a controlled engineering agent.

Concept diagram
Agent state
Memory + State
Tool-backed action
Durable conversation or workflow state that lets agents continue tasks safely across multiple steps.
Back to role-fit map
Agent Engineering#071

Subagents

Specialised child agents delegated bounded parts of a larger engineering or reasoning task.

Role lens

Reusable tools, MCP, state and delegation turn an LLM into a controlled engineering agent.

Concept diagram
Agent state
Subagents
Tool-backed action
Specialised child agents delegated bounded parts of a larger engineering or reasoning task.
Back to role-fit map
Agent Engineering#072

Skills

Reusable packaged instructions, tools and workflows that make AI-assisted engineering repeatable.

Role lens

Reusable tools, MCP, state and delegation turn an LLM into a controlled engineering agent.

Concept diagram
Agent state
Skills
Tool-backed action
Reusable packaged instructions, tools and workflows that make AI-assisted engineering repeatable.
Back to role-fit map
Agent Engineering#073

Hooks

Lifecycle automations that run before or after agent actions, tests, commits or deployment events.

Role lens

Reusable tools, MCP, state and delegation turn an LLM into a controlled engineering agent.

Concept diagram
Agent state
Hooks
Tool-backed action
Lifecycle automations that run before or after agent actions, tests, commits or deployment events.
Back to role-fit map
Safety + LLMOps#074

Guardrails

Deterministic checks that constrain model inputs, outputs or actions according to policy.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Guardrails
Controlled outcome
Deterministic checks that constrain model inputs, outputs or actions according to policy.
Back to role-fit map
Safety + LLMOps#075

Prompt Injection

Adversarial instructions hidden in user or retrieved content that try to override trusted system intent.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Prompt Injection
Controlled outcome
Adversarial instructions hidden in user or retrieved content that try to override trusted system intent.
Back to role-fit map
Safety + LLMOps#076

PII Redaction

Detecting and removing personally identifiable information before inappropriate exposure or model use.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
PII Redaction
Controlled outcome
Detecting and removing personally identifiable information before inappropriate exposure or model use.
Back to role-fit map
Safety + LLMOps#077

Secrets Management

Keeping API keys, credentials and tokens outside prompts and source code with controlled access.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Secrets Management
Controlled outcome
Keeping API keys, credentials and tokens outside prompts and source code with controlled access.
Back to role-fit map
Safety + LLMOps#078

Least Privilege

Giving users and agents only the minimum permissions required for the current task.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Least Privilege
Controlled outcome
Giving users and agents only the minimum permissions required for the current task.
Back to role-fit map
Safety + LLMOps#079

Audit Trail

A durable record of model, retrieval, tool, approval and deployment decisions for accountability.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Audit Trail
Controlled outcome
A durable record of model, retrieval, tool, approval and deployment decisions for accountability.
Back to role-fit map
Safety + LLMOps#080

Groundedness

The degree to which an AI answer is supported by authorised source evidence rather than unsupported generation.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Groundedness
Controlled outcome
The degree to which an AI answer is supported by authorised source evidence rather than unsupported generation.
Back to role-fit map
Safety + LLMOps#081

Hallucination

Unsupported or fabricated model output that must be reduced through retrieval, constraints and evaluation.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Hallucination
Controlled outcome
Unsupported or fabricated model output that must be reduced through retrieval, constraints and evaluation.
Back to role-fit map
Safety + LLMOps#082

Golden Dataset

A curated set of representative inputs and expected outcomes used to evaluate AI quality consistently.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Golden Dataset
Controlled outcome
A curated set of representative inputs and expected outcomes used to evaluate AI quality consistently.
Back to role-fit map
Safety + LLMOps#083

Regression Testing

Checking that prompt, model, retrieval or code changes do not break previously working behaviour.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Regression Testing
Controlled outcome
Checking that prompt, model, retrieval or code changes do not break previously working behaviour.
Back to role-fit map
Safety + LLMOps#084

Tracing

Following a request across model calls, retrieval, tools and services to diagnose behaviour and failures.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Tracing
Controlled outcome
Following a request across model calls, retrieval, tools and services to diagnose behaviour and failures.
Back to role-fit map
Safety + LLMOps#085

Latency

End-to-end response time, a key production constraint when choosing models, retrieval and tool paths.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Latency
Controlled outcome
End-to-end response time, a key production constraint when choosing models, retrieval and tool paths.
Back to role-fit map
Safety + LLMOps#086

Token Cost

The model-usage cost created by input and output tokens, context size and repeated calls.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Token Cost
Controlled outcome
The model-usage cost created by input and output tokens, context size and repeated calls.
Back to role-fit map
Safety + LLMOps#087

Model Routing

Selecting a model dynamically based on capability, cost, latency, policy or availability.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Model Routing
Controlled outcome
Selecting a model dynamically based on capability, cost, latency, policy or availability.
Back to role-fit map
Safety + LLMOps#088

Provider Fallbacks

Switching to an alternate model or provider when the preferred path is unavailable or unsuitable.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Provider Fallbacks
Controlled outcome
Switching to an alternate model or provider when the preferred path is unavailable or unsuitable.
Back to role-fit map
Safety + LLMOps#089

Retries

Controlled re-execution of transiently failed calls with limits, backoff and idempotency.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Retries
Controlled outcome
Controlled re-execution of transiently failed calls with limits, backoff and idempotency.
Back to role-fit map
Safety + LLMOps#090

Caching

Reusing safe prior results or computed context to reduce latency, cost and repeated work.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Caching
Controlled outcome
Reusing safe prior results or computed context to reduce latency, cost and repeated work.
Back to role-fit map
Safety + LLMOps#091

Idempotency

Designing repeated tool or API calls so retries do not create duplicate side effects.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Idempotency
Controlled outcome
Designing repeated tool or API calls so retries do not create duplicate side effects.
Back to role-fit map
Safety + LLMOps#092

API Contracts

Explicit request, response, error and version schemas that separate AI reasoning from deterministic services.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
API Contracts
Controlled outcome
Explicit request, response, error and version schemas that separate AI reasoning from deterministic services.
Back to role-fit map
Safety + LLMOps#093

Schema Validation

Checking structured inputs and outputs against types or schemas before downstream use.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
Schema Validation
Controlled outcome
Checking structured inputs and outputs against types or schemas before downstream use.
Back to role-fit map
Safety + LLMOps#094

OpenTelemetry

Open standard for collecting traces, metrics and logs across distributed production services.

Role lens

Production AI needs deterministic controls around probabilistic behaviour: security, traces, retries and cost.

Concept diagram
Runtime request
OpenTelemetry
Controlled outcome
Open standard for collecting traces, metrics and logs across distributed production services.
Back to role-fit map
Engineering Practice#095

Git

Version-control foundation for safe branching, review, traceability and rollback of software changes.

Role lens

AI-generated work should improve—not bypass—the review, testing, architecture and leadership disciplines of software engineering.

Concept diagram
Code change
Git
Quality + ownership
Version-control foundation for safe branching, review, traceability and rollback of software changes.
Back to role-fit map
Engineering Practice#096

Pull Requests

Review boundary for inspecting AI- or human-generated code before it enters protected branches.

Role lens

AI-generated work should improve—not bypass—the review, testing, architecture and leadership disciplines of software engineering.

Concept diagram
Code change
Pull Requests
Quality + ownership
Review boundary for inspecting AI- or human-generated code before it enters protected branches.
Back to role-fit map
Engineering Practice#097

Unit Testing

Fast tests for isolated functions and components used to catch deterministic software regressions.

Role lens

AI-generated work should improve—not bypass—the review, testing, architecture and leadership disciplines of software engineering.

Concept diagram
Code change
Unit Testing
Quality + ownership
Fast tests for isolated functions and components used to catch deterministic software regressions.
Back to role-fit map
Engineering Practice#098

Integration Testing

Tests that verify models, APIs, databases and services work together across real boundaries.

Role lens

AI-generated work should improve—not bypass—the review, testing, architecture and leadership disciplines of software engineering.

Concept diagram
Code change
Integration Testing
Quality + ownership
Tests that verify models, APIs, databases and services work together across real boundaries.
Back to role-fit map
Engineering Practice#099

Architecture Decision Records

Short versioned records explaining important architecture choices, trade-offs and consequences.

Role lens

AI-generated work should improve—not bypass—the review, testing, architecture and leadership disciplines of software engineering.

Concept diagram
Code change
Architecture Decision Records
Quality + ownership
Short versioned records explaining important architecture choices, trade-offs and consequences.
Back to role-fit map
Engineering Practice#100

Technical Leadership

Setting engineering direction while remaining close enough to implementation to unblock teams and raise standards.

Role lens

AI-generated work should improve—not bypass—the review, testing, architecture and leadership disciplines of software engineering.

Concept diagram
Code change
Technical Leadership
Quality + ownership
Setting engineering direction while remaining close enough to implementation to unblock teams and raise standards.
Back to role-fit map
Seen 1 / 100 · swipe to learn
Roll the dice