BrowseAll ecosystems
Everything in the atlas
16 of 2341
Domain: AI & Agents
Signals
Content: 2
More
Group by: Capability
EntryBySignalsLast activity
Patterns for using LLMs in ABAP developmentPatterns for using LLMs in ABAP developmentLars Hvam compares eight ABAP development setups: assistance in SE80 or Eclipse, virtual files and MCP over ADT, a centrally managed MCP service, sandboxes, local Git, Git with read-only system context, and temporary ABAP systems for development and CI. The overview describes where source changes, review and system access happen, including abapGit synchronization and abaplint checks. It is dated March 2026; its risk, cost and autonomy ratings are the writer’s assessments.
CAP in the age of AI #3 - from MCP to A2ACAP in the age of AI #3 - from MCP to A2AThe episode that finishes @cap-js/mcp with authorisation and then crosses to agents. @requires and @restrict are the same annotations CAP already has, and the plugin reads them to filter the tool list and the describe output per user and per request, so a model is not told about entities the caller cannot reach — where the OData metadata of the same service lists every exposed entity regardless. The second half annotates a service with @agent from @cap-js/agents, which serves it over A2A with an agent card and a chat preview, runs first against a mocked response and then against SAP AI Core bound with cds bind and a hybrid profile, and reuses the MCP adapter's tools rather than defining new ones.published 2026-08-21
CAP in the age of AI #4 - more on agentsCAP in the age of AI #4 - more on agentsAsking a human first, and telling the agent how to behave. Ordering a book through @cap-js/agents happens without a question being asked; the human-in-the-loop annotation on the action turns the same request into an approve-or-reject step, which the episode follows down to the A2A status update that stores the task in an input-required state in the database — A2A itself being JSON-RPC 2.0 over HTTP with server-sent events for the stream. The rest is instructions on disk: an agents.md in the service's own folder removes the question about how many copies to order, a skill under a skills folder is advertised by its description and read only when the agent decides the case applies, and either file makes the plugin build a deep agent instead of the plain one.published 2026-09-11
CAP in the age of AI #5 - agents wrap-upCAP in the age of AI #5 - agents wrap-upThe agents topic closed out on what sits around the agent rather than inside it, now that the documentation has moved into capire. The chat preview's session is persisted in the database rather than kept in memory, traces go out through @cap-js/telemetry to a self-hosted MLflow under the OpenTelemetry conventions for generative AI, and the panel is explicit that @agent reuses the MCP adapter's code programmatically without serving an MCP endpoint. It ends in the xtravels sample, where a travel agent delegates to a hotel-service sub-agent and reaches a separately started flight service over MCP through the service binding CAP resolves anyway. Earlier in the hour the same agent answers from a model running on the presenter's laptop, reached as an OpenAI-compatible endpoint.published 2026-10-02
Content-based Agents with SAP Joule Studio Agent BuilderContent-based Agents with SAP Joule Studio Agent BuilderJoule Studio, classic edition, the capability in SAP Build the page calls Joule Studio, is where an organisation creates and manages its own Joule skills and Joule agents; its Agent Builder is for a custom agent that handles a multi-step task by combining instructions, model-backed planning and reasoning, and a configured set of tools, Joule skills, document resources, MCP servers, or other agents run as subagents. The page states an agent executes inside the calling user's own security context through SAP Cloud Identity Services, and can be tried against a test conversation before deployment. That is the tool-and-turn-taking ground agent tooling covers, here built through Joule Studio rather than in application code.
Custom Joule SkillsCustom Joule SkillsA custom Joule skill is a workflow written to extend what Joule does out of the box for one specific business need, built and deployed through Joule Studio in SAP Build; the page states this is the supported way to extend Joule's built-in capabilities. Through Joule Studio a skill reaches SAP and external APIs via the destinations already configured in the BTP subaccount and Build environment, with the workflow itself versioned, monitored and deployable from the same tool.
Pro-Code AI AgentsPro-Code AI AgentsPro-code agents are the most flexible way the page frames building an agentic workflow on SAP BTP: full autonomy, custom API integrations and algorithms chosen by the developer, set against declarative alternatives such as content-based agents and Joule Scenarios. It names LangGraph as the framework it recommends for production, enterprise-grade agentic workflows, alongside the ReAct pattern as the plan-act-observe loop such an agent runs. That is the ground agent tooling covers, building an agent's own tool-calling loop rather than exposing an application to somebody else's agent.
Content FilteringContent FilteringContent filtering screens a large language model's input and output for low-quality, irrelevant or unsafe content: the four categories it names being hateful, violent, sexual and vulgar content, checking a request once on the way in and once on the way back. The page states smaller models can do this filtering efficiently, reading perplexity scores to flag anomalies, with the aim that SAP AI Core produces content meeting a stated safety standard, which is the ground content filtering covers on the BTP door.
Data MaskingData MaskingData masking replaces identifiable or confidential text in a prompt with placeholders such as <Person_0> or <Organization_1> before it reaches a large language model, using pattern matching and machine learning to find what to replace. The page states SAP AI Core supports it as a feature and sets out five steps for applying it, defining the masking scope, choosing a technique, preserving the prompt's intent, checking compliance and risk, then testing and monitoring the result. That is the shape data masking takes on the BTP door.
Access to Generative AI ModelsAccess to Generative AI ModelsSets out the plain way to call a deployed generative AI model on SAP BTP: a request that carries the prompt, the model and the parameters that control generation, with the same code able to switch between models. It lists data masking, prompt templating and filtering as features that come with the access rather than as code the caller writes, which is the shape LLM integration takes on the BTP door; the page's implementation sections sit behind its steps and were not read.
CAP in the age of AI #3 - from MCP to A2ACAP in the age of AI #3 - from MCP to A2AThe episode that finishes @cap-js/mcp with authorisation and then crosses to agents. @requires and @restrict are the same annotations CAP already has, and the plugin reads them to filter the tool list and the describe output per user and per request, so a model is not told about entities the caller cannot reach — where the OData metadata of the same service lists every exposed entity regardless. The second half annotates a service with @agent from @cap-js/agents, which serves it over A2A with an agent card and a chat preview, runs first against a mocked response and then against SAP AI Core bound with cds bind and a hybrid profile, and reuses the MCP adapter's tools rather than defining new ones.published 2026-08-21
CAP in the age of AI #5 - agents wrap-upCAP in the age of AI #5 - agents wrap-upThe agents topic closed out on what sits around the agent rather than inside it, now that the documentation has moved into capire. The chat preview's session is persisted in the database rather than kept in memory, traces go out through @cap-js/telemetry to a self-hosted MLflow under the OpenTelemetry conventions for generative AI, and the panel is explicit that @agent reuses the MCP adapter's code programmatically without serving an MCP endpoint. It ends in the xtravels sample, where a travel agent delegates to a hotel-service sub-agent and reaches a separately started flight service over MCP through the service binding CAP resolves anyway. Earlier in the hour the same agent answers from a model running on the presenter's laptop, reached as an OpenAI-compatible endpoint.published 2026-10-02
CAP in the age of AI #6 - vector embeddingsCAP in the age of AI #6 - vector embeddingsVector embeddings put to work in the incident management sample, answering whether this problem has been seen before. An embedding element of type Vector is calculated and stored from the vector_embedding function, so the database refreshes it whenever the text behind it changes, and cosine_similarity compares it with a freshly embedded string; because a title says little, a scheduled action first summarises each incident's conversation through a model on SAP AI Core and the embedding is taken from that summary. The similar-incidents list is a view joining incidents to themselves, with order and limit set in a before handler on the expanded query, and the vector element excluded from the service, since OData knows nothing about vectors — on SAP HANA Cloud the function is native, while SQLite needs packages added to the project.published 2026-10-09
CAP in the age of AI #7 - vector embeddings wrap-upCAP in the age of AI #7 - vector embeddings wrap-upThe same embeddings model taken from SQLite to SAP HANA Cloud, then handed to an agent. Locally it is @cap-js/ai that adds the vector functions SQLite lacks and downloads an all-MiniLM L6v2 ONNX model on first start; on HANA natural language processing has to be switched on, the model name is a mandatory argument, the vectors it returns are 768 dimensions, and dimensions must match on both sides of a comparison. HANA itself can be given SAP AI Core as a remote source for other embedding models, which a CDS.ql expression has to inject because the postfix notation only takes what the CDS model holds, and the hour ends with a search function over cosine similarity exposed over @cap-js/mcp, so an agent asked about low output from solar panels calls semantic search instead of writing the query itself.published 2026-10-16
Content FilteringContent FilteringContent filtering screens a large language model's input and output for low-quality, irrelevant or unsafe content: the four categories it names being hateful, violent, sexual and vulgar content, checking a request once on the way in and once on the way back. The page states smaller models can do this filtering efficiently, reading perplexity scores to flag anomalies, with the aim that SAP AI Core produces content meeting a stated safety standard, which is the ground content filtering covers on the BTP door.
Data MaskingData MaskingData masking replaces identifiable or confidential text in a prompt with placeholders such as <Person_0> or <Organization_1> before it reaches a large language model, using pattern matching and machine learning to find what to replace. The page states SAP AI Core supports it as a feature and sets out five steps for applying it, defining the masking scope, choosing a technique, preserving the prompt's intent, checking compliance and risk, then testing and monitoring the result. That is the shape data masking takes on the BTP door.
Document GroundingDocument GroundingDocument Grounding is a managed retrieval-augmented-generation service under the Generative AI Hub on BTP, which SAP's page on SAP AI Core places inside that service: it ingests documents from Amazon S3, SharePoint, SAP Document Management Service, SAP Work Zone or an SFTP server, and handles chunking, embedding and retrieval itself, refreshing content automatically when a source document changes. The page states this managed path is recommended for most enterprise scenarios because it covers the whole lifecycle end to end, and offers a separate Document Grounding Vector API for data that is already chunked and embedded, for cases that need finer control over chunking and metadata.
Prompt TemplatingPrompt TemplatingPrompt templating replaces a static prompt with a structured, reusable template whose placeholders are filled in at runtime, which the page says helps LLM responses stay consistent and lets the same instructions be adapted across tasks. On SAP AI Core this is implemented through the Prompt Registry, and the page states that storing prompts as plain strings in code is possible but not recommended. That is prompt management as a stored and managed object rather than a string embedded in the caller.
Prompt TemplatingPrompt TemplatingPrompt templating replaces a static prompt with a structured, reusable template whose placeholders are filled in at runtime, which the page says helps LLM responses stay consistent and lets the same instructions be adapted across tasks. On SAP AI Core this is implemented through the Prompt Registry, and the page states that storing prompts as plain strings in code is possible but not recommended. That is prompt management as a stored and managed object rather than a string embedded in the caller.
CAP in the age of AI #6 - vector embeddingsCAP in the age of AI #6 - vector embeddingsVector embeddings put to work in the incident management sample, answering whether this problem has been seen before. An embedding element of type Vector is calculated and stored from the vector_embedding function, so the database refreshes it whenever the text behind it changes, and cosine_similarity compares it with a freshly embedded string; because a title says little, a scheduled action first summarises each incident's conversation through a model on SAP AI Core and the embedding is taken from that summary. The similar-incidents list is a view joining incidents to themselves, with order and limit set in a before handler on the expanded query, and the vector element excluded from the service, since OData knows nothing about vectors — on SAP HANA Cloud the function is native, while SQLite needs packages added to the project.published 2026-10-09
CAP in the age of AI #7 - vector embeddings wrap-upCAP in the age of AI #7 - vector embeddings wrap-upThe same embeddings model taken from SQLite to SAP HANA Cloud, then handed to an agent. Locally it is @cap-js/ai that adds the vector functions SQLite lacks and downloads an all-MiniLM L6v2 ONNX model on first start; on HANA natural language processing has to be switched on, the model name is a mandatory argument, the vectors it returns are 768 dimensions, and dimensions must match on both sides of a comparison. HANA itself can be given SAP AI Core as a remote source for other embedding models, which a CDS.ql expression has to inject because the postfix notation only takes what the CDS model holds, and the hour ends with a search function over cosine similarity exposed over @cap-js/mcp, so an agent asked about low output from solar panels calls semantic search instead of writing the query itself.published 2026-10-16
Document GroundingDocument GroundingDocument Grounding is a managed retrieval-augmented-generation service under the Generative AI Hub on BTP, which SAP's page on SAP AI Core places inside that service: it ingests documents from Amazon S3, SharePoint, SAP Document Management Service, SAP Work Zone or an SFTP server, and handles chunking, embedding and retrieval itself, refreshing content automatically when a source document changes. The page states this managed path is recommended for most enterprise scenarios because it covers the whole lifecycle end to end, and offers a separate Document Grounding Vector API for data that is already chunked and embedded, for cases that need finer control over chunking and metadata.
Vector RAG vs Graph RAGVector RAG vs Graph RAGA section introduction rather than the comparison itself: it states that the choice between Vector RAG and Graph RAG depends on the domain and the task, and that Vector RAG is the one to use when the subject area has a clear description and the model must be held to specific data structures and attribute values. The four sub-pages it introduces, two on vector-based and two on graph-based RAG, carry the comparison. It sits on the same retrieval question retrieval-augmented generation answers, without naming CDS, CAP or a store.
Agentic AI Security & GovernanceAgentic AI Security & GovernanceNames seven threat categories agentic AI introduces beyond ordinary application security: prompt injection, excessive agency, supply-chain and model-integrity compromise, memory poisoning, improper output handling, resource abuse and agent-to-agent communication poisoning. Its answer is a defence-in-depth structure of security controls, input validation, privilege management, supply-chain checks, memory segmentation, output validation, execution quotas, secure communication, paired with governance and oversight: audit logging, explainability, human-in-the-loop review, fairness assessment and continuous monitoring. States the reason plainly: an agent decides, chains actions across turns, keeps memory between them and can act on real systems, a different risk profile from a single request-response call. That is the authorisation and audit-trail question tool governance is named for.