Adobe Experience Platform

The feature that automatically auto tags images in a document needs to be disabled for every user, or the tagging needs to be manual to eliminate obscene language from the tagging. I have an abstract image in my template that I send to clients. If I forget to disable the auto tagging, then the client will read a caption that suggests I am sending them a nude image. There is no figure in the graphic. There is no suggestion of a figure in the graphic, but Ai suggest that there is, and it confuses my clients. Even if it says that it is an “Ai generated caption”, my clients read it as something I, myself, inserted into their document.  I have turned it off. I have done it multiple times, successfully. However, I do not love that it is my sole responsibility to work around a program every single time I send out a document.  Please let me know if there is a way to disable this function so that I can forget about the fact it even exists. My clients are older and I have to explain to them ever

We use Apache CXF for SOAP and REST clients on AEM 6.5.2 LTS and install the necessary bundles in AEM. Currently we are forced to use Apache CXF 3.3.13, which was released in March 2022, because AEM still ships with org.apache.servicemix.specs.saaj-api-1.3, and newer versions of Apache CXF require saaj-api-1.4. There are a number of security vulnerabilities in Apache CXF that have since been patched since 2022 that we would like to remediate, see https://cxf.apache.org/security-advisories.html. Our goal would be to install Apache CXF 3.6.x, currently 3.6.10, as that is still actively supported, but the old SAAJ API is preventing us from doing that, see javax.xml.soap,version=[1.4,2) -- Cannot be resolved: Our ask is for AEM to upgrade to saaj-api-1.4, which would unblock us from upgrading Apache CXF.I can provide a full list of bundle JARs that we install related to Apache CXF if needed.

There should be a setting in the new launch to be able to track s.products without going to the custom code to build the product string.Kind of like how the old GTM had a checkbox to enable e-commerce tracking and it would just read the dataLayer Enhanced Ecommerce (UA) Developer Guide  |  Google Tag Manager for Web Tracking  |  Google Developers

DescriptionI wish there were a more comprehensive panel for Engageable Profiles in Adobe Experience Platform. Why is this feature important to youPrimarily, it would provide organizations with a deeper understanding of profile usage.Recently, our organization faced an issue that resulted in nearly one million profiles becoming engageable due to an error. Identifying the root cause required us to use Query Service to identify the journey, then open the journey to determine the associated audience, and finally analyze the audience to understand how those profiles were ingested.This investigation process was very time-consuming. Having an out-of-the-box monitoring capability for Engageable Profiles would significantly simplify troubleshooting and usage analysis. Important monitoring could include:New engageable profiles by day Which journeys are triggering new profile engagement, allowing users to display a sample of the newly engaged profiles Which namespaces are being used by those jour

Problem Statement :How can we support near real-time or frequent audience evaluation in AEP, when the platform only allows one scheduled batch segmentation run per day and does not support sub-daily automation? Solution: In Adobe Experience Platform (AEP), batch segmentation is designed for daily execution (one run per day per sandbox). This becomes a challenge when business requires more frequent audience refresh (hourly or near real-time use cases).To handle this, AEP provides a combination of on-demand execution, optimization strategies, and alternative approaches. 1. Understand the Core LimitationOnly 1 automated batch run per day per sandbox No hourly or multiple recurring schedules supported Schedule timing (default 5:00 AM UTC) can be modified only via API, not UI👉 This means native batch segmentation cannot meet high-frequency use cases 2. Use Flexible Audience Evaluation (FAE) – On-Demand RunsTo overcome this limitation, AEP provides Flexible Audience Evaluation (FAE).✅ What

Have you ever noticed a strange behavior in Adobe Experience Platform (AEP)? You can build a structural blueprint called a Schema with a special map called an identityMap. You can add an email and a phone number as normal identities, but leave the "Primary Identity" field completely blank. The system still allows you to click "Enable for Profile" and save it without any errors at all!It looks like everything is working perfectly. But underneath the hood of Real-Time CDP, this creates a hidden trap: the system will connect your IDs in the map, but it will fail to create actual customer profiles.Let’s understand why this happens using a simple story based on a real-world company: Walmart. The Problem: Missing Profiles Imagine Walmart is tracking digital shopping activity using an AEP Schema. A customer goes online and makes a purchase, providing their phone number and email.Because the Walmart Schema has an identityMap at its root, AEP stores this transaction. Later, a data analyst check

Enhancement SummaryRequesting a product enhancement in Adobe Tags to support the ability to transfer, copy, or promote selected components from one library into another library within the same property. This capability would allow users to move changes such as data elements, rules, extensions, and related configurations between libraries without manually recreating or porting them.Current StateAdobe Tags currently allows teams to create and manage libraries as part of the publishing workflow. However, when changes are created in one library, there does not appear to be a native way to push or transfer those changes into another existing library.As a result, teams are required to manually recreate or port data elements, rules, and other configuration changes from one library to another. Business ProblemThe current manual process introduces unnecessary operational risk and inefficiency.Manual porting between libraries can result in:Missed data elements or rules Inconsistent configuration

we have +60 languages to select but many times we just need some of them and we always know it is around 22 as exampleI could be really cool with a kind of counter of selected language, to help us verify if we have selected the right amount of languages  

Many enterprise Adobe Commerce customers are interested in modernization, but a full immediate transition to composable or ACCS is often difficult because of existing customizations, integrations, operational dependencies, and business risk.A possible approach could be an officially documented “progressive modernization blueprint” that helps organizations move step-by-step instead of forcing a large transformation program.The idea would be to provide guidance around:introducing App Builder gradually alongside existing PaaS implementations identifying which customizations should move first to extensibility services preparing APIs and event-driven integrations early coexistence strategy between legacy modules and composable services phased operational governance and monitoring ACCS readiness assessment framework recommended migration sequence for enterprise implementationsThis could help customers:reduce migration risk modernize incrementally improve long-term maintainability align inves

Your storefront now has two audiences. Most architectures are built for only one.Adobe summit 2026 made this crystal clear: AI agents are actively shopping on behalf of consumers and most storefronts aren't built for them.During the 2025 holiday season, generative AI tools drove a 693% increase in traffic to retail sites year-over-year. And those AI-referred shoppers convert 31% higher and generate 254% more revenue per visit.The two audiences your storefront must now serve:HUMAN SHOPPERS (you know this one):Browse pages, use search, click through checkoutRespond to promotions and personalized recommendationsUX, Core Web Vitals, and page performance still matterAI AGENTS (new requirement):Query your catalog via API : they never see your homepageCompare products across multiple LLMs on behalf of usersExecute purchases autonomously using protocols like UCP and ACPNeed machine-readable product data: rich descriptions, structured specs, real-time pricing and inventoryWhat this means archit

Your ecommerce platform isn't slow because of traffic. It's slow because of architecture.Global ecommerce is heading toward $7.4 trillion by 2026. And the brands eating that market share all have one thing in common: they ditched monolithic platforms years ago.As someone who architects large-scale commerce platforms daily, Adobe Commerce, Shopify Plus, headless builds, here's what I'm seeing on the ground:The shift to Composable Commerce is no longer a trend. It's the baseline.What composable actually means in practice:→ Your frontend (Next.js / React) is completely decoupled from your backend→ Your commerce engine handles cart, catalog, pricing, nothing more→ AI and personalization run as a dedicated layer, not bolted on plugins→ ERP, CRM, PIM, OMS connect via event-driven APIs not fragile point-to-point integrationsAnd the new frontier nobody's talking about enough: Zero-Click Commerce.AI agents are starting to shop on behalf of consumers. They don't browse. They decide. If your prod

Description - the current Web SDK "set-consent" behaves a little unexpected in different situations, leading to unwanted results, whereas a the legacy Visitor ID service behaved differently.   Examples:   A) A new visitor denies cookies. With the Visitor ID service, you would've just denied the consent categories, mapped to the Adobe tools without having any whatsoever cookie written. With the Web SDK however, if you send a general "opt out" through the "set-consent" action, two things happen "kndctr" cookies are set (!) if the company has a RT-CDP license, a pseudonymous profile is created (correct me if am wrong!) with the consent preferences attached to it   In my opinion, none of the two above make sense, and cookies written lead to potential legal problems.   Now, I am a seasoned Launch developer and have a workaround that involves local storage and some other constructs to make sure the "set-consent" rule only triggers (and writes cookies and creates a profi

Description - Many times database connection on Query Service get disconnected due to various reasons (limit of users reached being one of it). This affects the focus work as it disrupts the query execution, to avoid it User should have a connection refresh button which will allow the Database connection to be re-established without having to refresh the entire browser multiple times.Why is this feature important to you - Helps users to continue important focus query time without disruptionsHow would you like the feature to work - A simple refresh button the page should reestablish the connection to the database.Current Behaviour - Users need to copy the entire query and refresh the browser, it clear all the query results previous queries might have returned on the browser.

We have a legal/compliance requirement to store all our iterations/versions of our documents. In addition to retaining them for an indefinite period, we need to be able to view the various iterations/versions & markups (created via Adobe Acrobat) that have been created during the document creation & approval process of the document. I loaded a couple iterations of a document that had 7 iterations/versions created in another content management tool (again, markups created via Adobe Acrobat) and tried to access the prior iterations in the timeline preview but do not see the markups in the preview. I double-clicked on the current document and do see there are markups, but you can’t select them or hover over them to see the actual markups. We need an easy way to view all iterations/versions of a document and be able to see the markups made from version to version. 

I am an Adobe and (now also) ai consultant in the fashion industry for over twenty years. I am absolutely stoked about Firefly boards and teach all designers i meet to work on it. There are two potential features that would make Firefly boards unbeatable. 1: The possibility to add style-guide information on a specific board or maybe a ‘project folder would be even better’. This will make Firefly boards instantly the number one ai tool for brands (not only fashion). 2: Integrated LLM of choice. I now teach designers to use their favorite llm to create detailed prompts for generating images and video on a firefly board. How nice would it be if there would be a sidebar with an integrated LLM. Even better if the user can choose their own favorite llm (please include Mistral for the European market).  

In multi-brand Adobe Experience Platform implementations, consent information is often captured as event data, with each brand emitting its own consent events.These consent signals are then elevated to the profile level using Computed Attributes, so they can be reused consistently across audiences, consent policies, and activation use cases.However, Computed Attributes currently have a maximum lookback period of 6 months. If no qualifying events are present within that window, the computed attribute value is nullified, even when no explicit consent change or withdrawal has occurred.This behavior creates challenges when modeling consent, which is inherently a stateful attribute, not a time-bound behavioral signal. Suggested EnhancementProvide an option for Computed Attributes to retain the last known value beyond the lookback window, unless:A new qualifying event updates the value, orAn explicit revocation / change event is received.This could be implemented as:A configuration option su

With companies becoming more security-aware and device manufacturers blocking requests to certain domains, people are looking for opportunities to bring their marketing tools in a first-party context. With Launch, we can either host the library on Adobe's domain or our own webserver, but not behind a transparent reverse-proxy. After some experimentation, I have found that the individual extensions and files are still loaded from Adobe's domain, even if the initial Launch library is hidden behind a proxy. Conceptually it should be an easy addition to allow users to hide the adobedtm Domain behind their company's proxy. From a feature perspective, I would like to specify both a domain and a path from where Launch should attempt to load files instead of the standard Adobe domain and path. The files should still be pushed to Akamai and be available there, but should not reference files from the Adobe domain. This would make it significantly easier for companies like my own to bring La

Currently, FIFO is not guaranteed when ingesting data to Profile, even if you have a timestamp available. This has been confirmed, both theoretically and practically, by external Adobe consultants. My feature request is to have the option to enforce FIFO upon profile ingestion. It could be a toggle, that would be off by default, when configuration Base connection and/or dataflows.  It will, potentially, have processing time implication, but from my business standpoint, I would rather have correctly updated profiles, a bit slower, than fast updates that are potentially incorrect. This would be a massive product improvement in terms of functionality, personalization and data activation in general. 

Currently, it seems that even though the data prep UI suggests this is possible, that mapping a nested data structure to a target field as a calculated field is not supported and ends up in a stalled data ingestion. Example: I have a nested source data structure, from which I want to extract a specific email address from index 0 of an array. This will always be the same index.Source dataNow, when I try to use data prep functions, the green checkmark appears, as well as the desired email address in the preview.Real mapping exampleSo far, everything looks as it should work. Until you start the workflow and it never ends.Now, I was 99% done creating my support ticket, when I read the AI-suggested ticket solutionAI suggested response.Possible solutionsbest case: Adobe integrates support for this deterministic path.  acceptable case: fix the UI to not show the false positive and maybe instead point out the issue. worst case: nothing changes, and I won’t be the last to scratch his head about

Description - These days in Launch's Web SDK extension, it is not possible to manually override the "_id" attribute in Web SDK XDM object data elements. As a workaround, if one wanted to manually define this _id, one would have to code an alloy "sendEvent" and specify the _id there. The approach / idea resembles very much the hybrid tracking approach for client- and server-side, that many 3rd party vendors suggest these days, where a deduplication key can be specified that either drops or merges events with duplicate ids.   Why is this feature important to you Completeness of data and fastest possible data availability: Web SDK data relies on cookie consent which may not have been given by the users i.e., an order event may or may not be captured by the Web SDK this lack of order data significantly impacts the number of addressable members in an audience, that may be built upon CRM and commerce data while this commerce data can also be streamed into the Platform from the commer

Description - as soon as an Audience is created that is using aggregation e.g., to evaluate a certain number of page views with a given property like a content category, the only working evaluation is "Batch" which is obviously not suitable if you want to do next page personalization based upon a user's content affinity.  Why is this feature important to you - It seems to be a standard feature in nowadays' CDPs to be able to define some kind of real-time "Badges" based upon user behaviour which can include aggregated values.This would be exactly that. Audiences on the Edge should support this as well since Batch evaluation is just too slow for next-page personalization. How would you like the feature to work - Allow aggregation on Edge or Streaming Audiences to be closer to real-time Current Behaviour - Upon save, an error is shown and none of the available segmentation methods works except for "Batch"    

Currently, if you share an audience to e.g. Facebook Custom Audience, you can do only a 1:1 mapping of your xdm schema field. As some destinations have specific requirements, it would be great to have like a calculated field as an option (same as in datastream mapping). One example would be to lowercase your xdm schema fields, as e.g. Facebook Custom Audience needs firstname / lastname in lowercase. As of now, I have to ingest an additional field with lowercase firstname / lastname to make it happen (or do some query service magic). I guess some simple adjustments / transformations of your current xdm schema fields would be very helpful.

Description - Currently, audiences defined in Adobe Experience Platform (AEP) are not directly available in Customer Journey Analytics (CJA). This limits the ability to perform advanced segmentation and analysis within navigational reports.his functionality was previously available when using Adobe Audience Manager with Adobe Analytics, where the dimension "Audience Name" allowed direct access to existing audiences. This enabled the kind of segmentation and analysis that is now missing between AEP and CJA.Why is this feature important to you - The lack of direct integration prevents:a)Analyzing navigational behavior of users depending on specific AEP audiences (designed with attribute conditions).b)Enriching navigational reports with segmentation criteria defined in AEP, especially when such data is not available through connected dataviews in CJA; as this attribute data is not useful for daily reports in CJA.This affects the efficiency of analysis and the ability to make bus

We have got ourselves into a situation where we have made many revisions to rules/elements which we have not published and no longer intend to publish. What is live and what is in our elements/rules are very different and it is now very difficult to get back to a clean baseline within rules and elements. The only way to do this is to manually revert every changed rule/element to match the current live version - very time consuming and prone to error. Please can you consider introducing a way to quickly update multiple elements/rules to match the live version?  

  Description -While working in the source workspace, I've noticed two key usability gaps that affect data management efficiency:1. Missing Date-Based Sorting/Filtering.Currently there is no option to filter or sort files on a source folder by their date of availability or uploaded time.This make it difficult to locate newly available files or track which data arrived most recently -especially when the number of files is large.Adding a simple date filter or filter (e.g: Last 7 days, 30 days or custom range) would make it much easier to ingest files or validate latest ingestions.2. Dataflow Filter Performance IssuesThe existing filter under Dataflows is slow to load and often returns incomplete or inaccurate results.Optimizing the Dataflow filter logic and improving response times would significantly help teams manage multiple dataflows and save time.Why is this feature important to you -Easier identification of newly uploaded data to ingest.Faster and more accurate Dataflow m