Shape the future of IBM watsonx Orchestrate

Start by searching and reviewing ideas others have posted, and add a comment (private if needed), vote, or subscribe to updates on them if they matter to you.

If you can't find what you are looking for, create a new idea:

  1. stick to one feature enhancement per idea

  2. add as much detail as possible, including use-case, examples & screenshots (put anything confidential in Hidden details field or a private comment)

  3. Explain business impact and timeline of project being affected

[For IBMers] Add customer/project name, details & timeline in Hidden details field or a private comment (only visible to you and the IBM product team).

This all helps to scope and prioritize your idea among many other good ones. Thank you for your feedback!

Specific links you will want to bookmark for future use
Learn more about IBM watsonx Orchestrate - Use this site to find out additional information and details about the product.
Welcome to the IBM Ideas Portal (https://www.ibm.com/ideas) - Use this site to find out additional information and details about the IBM Ideas process and statuses.
IBM Unified Ideas Portal (https://ideas.ibm.com) - Use this site to view all of your ideas, create new ideas for any IBM product, or search for ideas across all of IBM.
ideasibm@us.ibm.com - Use this email to suggest enhancements to the Ideas process or request help from IBM for submitting your Ideas.

ADD IDEA

All ideas

    Enable tools to use member credentials through Microsoft Teams and the Orchestrate Agent API

    Requested enhancement Enable tools to use member credentials when an agent is accessed through external channels. IBM watsonx Orchestrate should identify the requesting user and run tool actions on behalf of that user, using the credentials and pe...
    6 days ago in  0 Submitted

    Create CCA MCP Server builder using workflows UI.

    Banco do Brasil (BdB) is one of the world's largest Watson Assistant users and is actively migrating their customer-facing virtual assistant to watsonx Orchestrate (wxO) using the Customer Care Agent (CCA) model. Building CCA solutions today requi...
    20 days ago in  0 Submitted

    Per-sentence source-span grounding flags

    Problem: In document-grounded QA, the dominant failure isn't invented facts but interpretive overconfidence — unsupported characterizations of sources laundered into confident general statements. Fluency-optimized models can't supply claim-level s...
    about 12 hours ago in  0 Submitted

    Confidence-reset wrapper after observed failures

    Problem: LLMs grow more confident even after failing — they lack the post-task error signal humans use to recalibrate. An agent that just failed badly will confidently fail again. Idea: Add an Orchestrate "confidence reset" wrapper that tracks obs...
    about 12 hours ago in  0 Submitted

    Calibrated uncertainty phrasing with verbosity penalty

    Problem: People systematically overestimate what LLMs get right, and longer explanations inflate trust without improving accuracy — verbosity masquerades as reliability. Agents persuade when they should calibrate. Idea: Add an Orchestrate response...
    about 12 hours ago in  0 Submitted

    Anti-underaction nudge when clinical agents abstain

    Problem: Hiding uncertain AI outputs (selective prediction) fixes automation bias but creates a new harm: clinicians told "the AI abstained" underdiagnose (+18% missed diagnoses) and undertreat (+35% missed treatments). Abstention changes the erro...
    about 12 hours ago in  0 Submitted

    Assumptions-audit checklist before plan approval

    Problem: As agents present multi-step plans, humans over-rely on the plan itself — and common "critical thinking" nudges like what-if prompts feel helpful but don't change behavior while adding cognitive load. Idea: Add an Orchestrate plan-review ...
    about 12 hours ago in  0 Submitted

    Uncertainty head for skewed enterprise data

    Problem: Standard uncertainty methods assume clean, balanced data. Enterprise data is long-tailed and noisy, and rigid uncertainty models give badly calibrated "I don't know" signals exactly where agents face the messiest inputs. Idea: Add a drop-...
    about 12 hours ago in  0 Submitted

    Tight adaptive abstention sets for human tie-breaking

    Problem: When agents abstain or hedge, they produce over-wide candidate lists (useless) or overconfident single answers (dangerous). Honest uncertainty should be tight and adaptive, not vague. Idea: Add an Orchestrate abstention module fusing evid...
    about 12 hours ago in  0 Submitted

    Evidential-uncertainty flagging for medical-imaging agents

    Problem: Medical-imaging AI fails dangerously exactly where it is most confident-looking: large segmentation errors hide behind plausible outputs, and today's uncertainty scores (entropy, dropout) correlate poorly with actual mistakes. Idea: Add a...
    about 12 hours ago in  0 Submitted

    By clicking the "Post Comment" or "Add Idea" button, you are agreeing to the IBM Ideas Portal Terms of Use.
    Do not include IBM confidential, company confidential, or personal information in any field.
    Having problems accessing this portal? Describe the problems in an email to ideasibm@us.ibm.com.