Tools by product. Filter, or search by what you want to do.

• Support QARead
  query_coach_tasks

  Pulls the coaching queue sorted by lowest triggering Quality Score, filtered by status, manager, agent, date window or SLA cutoff. Answers which agents need a conversation this week and who is overdue.

• Support QARead
  get_coach_task

  Opens one coaching item end to end: the manager and agent, what triggered it, the analysis window, the due date and whether the check-in happened. Read it before the coaching conversation starts.

• Support QARead
  query_coach_sessions

  Shows the manager-led sessions that were actually logged, filterable by manager and date window. Set it beside the queue to see which assigned coaching turned into a real conversation.

• Support QARead
  list_coach_rubric

  Returns the scored behaviors your coach and analyst briefs grade against, each with its allowed values and the ones that count as failing. The place to learn what good looks like for your team.

• Support QARead
  get_coach_adherence

  Reports per manager how much assigned coaching got done: tasks created, completed, open and expired, sessions held, completion rates and coverage, with a roster rollup and optional trend buckets.

• Support QARead
  query_disputes

  Surfaces the times a person challenged an AI quality tag, filtered by status, agent, manager, tag and date. Shows where the AI and your team disagree, and which tags keep getting pushed back.

• Support QARead
  query_review_queue

  Lists the AI tags the model was unsure about and flagged for a human check. Shows how big the pile is and which tags it clusters in.

• Support QARead
  get_review_item

  Opens one low-confidence item and shows, for each tag, the value the AI chose, how sure it was and the evidence from the conversation behind that call.

• Support QARead
  query_audits

  Lists the manual QA audits your reviewers completed, with calibration stats on how often the AI's tags match their human verdicts.

• Support QARead
  query_audit_trends

  Charts agreement between the AI and human auditors over time, including how many tags were upheld versus corrected. Answers whether the machine is getting more or less trustworthy.

• Support QARead
  query_breaches

  Lists compliance and policy breaches filtered by status and severity, so risk and QA leads can see what is still open and what is serious.

• Support QARead
  get_coach_impact

  Measures whether coaching works: quality scores before and after a session, compared against agents who were not coached. Answers whether the coaching investment is moving the numbers.

• Support QARead
  query_coach_verifications

  Lists the coaching check-ins that are due and still unverified, worst-scoring agents first. Use it to chase the follow-ups that confirm coaching changed behavior.

• Support QARead
  get_analyst_brief

  Returns the Support QA analyst brief for the whole tenant, one line of business or a single manager: a written readout of quality performance you can hand straight to a leader.

• Support QARead
  get_bot_analyst_brief

  Returns the analyst brief on bot performance for the whole tenant or one line of business: a written readout of how your AI agents are doing, separate from the human team.

• MonitorRead
  query_feedback

  Searches every connected channel by meaning or exact keyword and returns normalized items in the customers' own words. Use it to read what people said about a launch, an error code or a feature.

• MonitorRead
  count_feedback

  Returns just the number of items matching a search or filter combination, plus a link to the same view in unitQ. Fast volume checks, such as how many reports an error code drew this week.

• MonitorRead
  query_feedback_summary

  Gives a short AI narrative of what a topic is about, the exact total, a handful of linked quotes and the categories it concentrates in. A one-call probe before you dig deeper.

• MonitorRead
  query_feedback_subtopics

  Clusters matching items with AI to reveal themes your taxonomy does not name yet, with estimated sizes. Shows what people are really complaining about inside a broad category.

• MonitorRead
  query_feedback_insights

  Returns unitQ's own AI findings, newest first: opportunities, risks and spikes, each with likely cause, who is affected, impact and the evidence quotes behind it. Filter by audience or date.

• MonitorRead
  query_feedback_categories

  Ranks the categories in your taxonomy by volume, period-over-period change or impact, with optional breakdowns and timeseries. Answers which issues are growing fastest and which ones matter most.

  Rank issues by impact
• MonitorRead
  list_filters

  Lists the dimensions you can slice by, such as platform, with plain-language descriptions. Call it first to learn the exact names to use in every other query.

• MonitorRead
  query_filters

  Shows the values inside a dimension and how volume splits across them, for example iOS versus Android versus Web. Typo-tolerant matching finds the right value even when you misspell it.

• MonitorRead
  list_metrics

  Shows every measure you can chart, with definitions, value types and ranges. Includes unitQ Score, Net Sentiment, Average Rating and volume, plus any your workspace defines.

• MonitorRead
  query_metrics

  Returns one metric as a single number or a timeseries, optionally broken down by any dimension and scoped to a search. Answers how sentiment or unitQ Score moved after a release, and for whom.

  Rank issues by impact
• MonitorRead
  query_cohort_metrics

  Defines a group of users by one set of criteria, then measures a metric across everything they said in a separate window. Example: how sentiment looks now for users who reported a payment dispute last quarter.

  Rank issues by impact
• MonitorRead
  query_cohort_categories

  Ranks issue categories across all the input from a defined group of users. Answers what else the people who hit login errors last month complain about.

  Rank issues by impact
• MonitorRead
  query_cohort_relative_metrics

  Aligns each user's series to their own first touch and measures a metric on a day-1, day-2, day-3 axis. Built for decay, retention and follow-up curves instead of calendar dates.

  Rank issues by impact
• MonitorRead
  query_dashboards

  Finds saved dashboards by keyword and returns each one's name, description and charts. Use it to locate "my exec dashboard" before you pull the numbers behind it.

• MonitorRead
  get_dashboard_data

  Pulls the full data behind every chart on one saved dashboard, with optional filter overrides or a single date range applied to all charts. Reproduces what you see in the app.

• MonitorRead
  query_alerts

  Shows which alerts are firing right now, or replays alert events over a date range, with status, trigger date and how much volume tripped each one. Defaults to critical and warning.

• MonitorWrite
  create_feedback_set

  Saves a scope as a reusable handle: pick subtopic clusters, explicit item IDs or a filter, then pass the returned set ID to the read tools to keep every question on the same slice.

• ResearchRead
  list_interviews

  Lists every study in the workspace, optionally only the published ones. Start here to find the study ID that every other research tool asks for.

• ResearchRead
  get_interview

  Returns the full definition of one study: its questions, follow-ups and settings. Read it before editing so changes keep what already works.

• ResearchRead
  list_responses

  Returns the respondent submissions for a study, newest first, as lightweight rows with outcome and score. Scan the list, then open the ones worth reading in full.

• ResearchRead
  get_response

  Opens one respondent's transcript, either truncated per turn for scanning or complete for analysis. Shows what the person actually said, word for word.

• ResearchRead
  get_response_analysis

  Returns the saved AI analysis of one respondent: summary, highlights, lowlights and learnings. Reads the cached result instead of running a new analysis.

• ResearchRead
  get_aggregate_report

  Returns the latest cross-respondent report for a study: themes, sentiment, per-question breakdown, recommendations and an executive summary. Reads the saved version, so it comes back instantly.

• ResearchRead
  ask_responses

  Answers a plain-language question about a study's collected answers in a few sentences, with every point cited to a verbatim respondent quote. Filters and follow-up turns narrow what it reads.

• ResearchRead
  translate_questions

  Translates a study's title, intro, questions, follow-ups, options and scale labels into another language and hands back the result unsaved, so you can review it before applying.

• ResearchRead
  parse_questions_from_file

  Reads a CSV, DOCX or PDF and drafts a study from it: title, description and questions, ready to pass to the create step. Turns an existing questionnaire into an interview.

• ResearchRead
  get_share_url

  Returns the public link respondents use to take a study, plus a template showing where to add URL parameters for panels, mail merges or internal tools.

• ResearchRead
  list_voices

  Lists the voices a voice-mode interview can speak with, each with an ID you pass when creating or editing a study.

• ResearchRead
  list_automations

  Lists the closed-loop rules watching a study's completed responses, with their conditions and actions. Read it before adding a rule, or to find out why an alert stayed silent.

• ResearchRead
  list_shared_reports

  Lists the share links created for a study's report, with access counts, whether each link is active and whether it updates as responses arrive.

• ResearchRead
  list_avatars

  Lists the AI avatars in your workspace with their name, voice and build status. Avatars host the video-call interviews, so this shows which ones are ready to use.

• ResearchWrite
  create_interview

  Creates a study from a brief and returns its ID as an unpublished draft. Covers questions, voice or chat mode, branding, welcome and thank-you messages, so a research plan becomes a study in one call.

• ResearchWrite
  duplicate_interview

  Clones a study into the same workspace as an unpublished copy. Use it to run a variation, such as new wording or another audience, without touching the original.

• ResearchWrite
  update_interview

  Edits only the fields you pass: replace the questions, or adjust settings such as branding and endings. Fetch the study first when you want to tweak a single setting.

• ResearchWrite
  publish_interview

  Opens a study to respondents once it has at least one question and every question has text. Run it only when someone has explicitly asked to launch.

  Can reach respondents.
• ResearchWrite
  unpublish_interview

  Takes a study offline so respondents can no longer open it. Use it to close a study without deleting anything.

• ResearchWrite
  generate_response_analysis

  Runs the AI analysis on one respondent's transcript, producing a summary with highlights, lowlights and learnings. Takes 10 to 30 seconds, then stays cached for reuse.

• ResearchWrite
  generate_aggregate_report

  Runs the cross-respondent analysis for a study: themes, sentiment, per-question breakdown, recommendations and an executive summary. Takes up to a minute on large studies, then stays cached.

• ResearchWrite
  create_automation

  Sets a closed-loop rule on a study: when a completed response matches your conditions, alert the study's followers, email the respondent or send a signed webhook. Includes a ready detractor alert.

  Can reach respondents.
• ResearchWrite
  update_automation

  Renames a rule, switches it on or off, or replaces its conditions or actions. Replacements swap the whole list, so read the existing rule first and send everything you want to keep.

  Can reach respondents.
• ResearchWrite
  delete_automation

  Permanently removes a rule and its delivery history. It stops acting on future responses and leaves recorded ones alone. Switch it off instead when you may want it back.

• ResearchWrite
  create_shared_report

  Creates a public link to a study's aggregate report, with an optional expiry. Hands stakeholders the findings through a single link.

• ResearchWrite
  create_avatar

  Builds an AI avatar from a photo, with a name, personality and opening line chosen for you when omitted. It takes a few minutes to process, then hosts video-call interviews.

• ResearchWrite
  update_avatar

  Changes only the fields you pass, such as its knowledge documents. A new photo sends the avatar back for a few minutes of processing; other edits apply immediately.

• ResearchWrite
  delete_interview

  Permanently removes a study together with every response and analysis collected under it. There is no undo, so confirm the target first.

  Irreversible. Deletes its responses too.
• ResearchWrite
  delete_response

  Permanently removes one respondent's submission and transcript from a study, for example a test run. There is no undo, so check the ID first.

  Irreversible.
• CompeteRead
  get_benchmark

  Ranks every app in the workspace by unitQ Score, defined as 100 minus the percent of quality issues, with the previous period and review counts. The headline view of where you stand.

• CompeteRead
  get_theme_matrix

  Counts tagged mentions per topic and app, split positive and negative, with bug counts. Shows where each app wins and loses, grouped by theme.

• CompeteRead
  get_reviews

  Searches raw reviews by text, app, star rating, sentiment, topic or the brand named, with filters for bugs and feature requests. Returns 25 at a time with the total.

• CompeteRead
  get_topic_quotes

  Returns representative verbatim reviews for one topic from the theme matrix, with sentiment. The evidence behind a single cell.

• CompeteRead
  get_mention_quotes

  Returns recent rival reviews that name a given brand, verbatim, drawn from reviews, Reddit and Trustpilot. Hear exactly how customers talk about the competition.

• CompeteRead
  get_score_series

  Plots weekly unitQ Score for every app over time. Compare weeks and find the inflection point behind a drop before you drill into topics.

• CompeteRead
  get_ratings

  Summarizes stars per app: the average, the 1 to 5 distribution and the share of reviews that are bugs or feature requests.

• CompeteRead
  get_rating_drivers

  Lists the topics tagged on reviews within a star band, showing what drives 1-star anger or holds an app back from 5 stars.

• CompeteRead
  get_cross_mentions

  Counts how often reviews of one app name-drop another brand, pair by pair. Shows who gets mentioned, and where.

• CompeteRead
  get_mention_themes

  Synthesizes what those name-drops mean, such as switching intent, into themes for each brand, with cited quotes behind every theme.

• CompeteRead
  get_kano

  Classifies each topic by its star-rating impact when mentioned negatively, positively or not at all. Separates must-haves from delighters.

• CompeteRead
  get_insights

  Returns the generated opportunity cards, framed as attack, gap or shift, each backed by evidence quotes from real reviews.

• CompeteRead
  get_strategy_briefs

  Returns the Porter five forces and Blue Ocean ERRC briefs computed for your workspace, ready to drop into a strategy discussion.

• CompeteRead
  get_weekly_report

  Returns the data snapshot behind the latest weekly competitive brief: the scoreboard, the biggest movers and the quote of the week.

• CompeteRead
  get_news_board

  Shows press coverage for your peer set in three columns: official newsroom stories, earned coverage, and stories that map to a topic reviews already discuss. Reprints collapse into one story with an outlet count.

• CompeteRead
  get_news_column

  Pages through one column of the news board, for the whole peer set or a single app, over a date window, with reprint and outlet counts on every row.

• CompeteRead
  get_news_story

  Opens one story: headline, publish date, every outlet that carried it and, when one exists, the review topic it overlaps with plus a negative quote. A topical overlap, never causation.

• SocialRead
  read_inbox

  Reads the newest customer conversations from DMs, mentions and reviews, with their messages and any pending AI draft. Filter by assignee or status; the response carries the exact match count.

• SocialRead
  read_conversation

  Opens one full thread: every message plus any pending AI draft. Use it to see the whole exchange before deciding what to do next.

• SocialRead
  search_conversations

  Full-text search across customer names, handles and message text. Finds every thread where someone raised a given product, outage or person.

• SocialRead
  read_analytics

  Returns the current KPI snapshot: volumes, sentiment split, response times and channel breakdown. The quick answer to how the social desk is doing.

• SocialRead
  read_analytics_series

  Returns sentiment and volume trends over time, so you can see when a conversation turned and how fast it grew.

• SocialRead
  read_alerts

  Lists current crisis alerts: conversations scored by severity that need attention first. The triage view when something is brewing.

• SocialRead
  read_mentions

  Reads public posts and comments that reference your brand, from social listening. Shows what is being said outside your own inbox.

• SocialRead
  read_influencers

  Lists the influencers you track along with their metrics, showing who is driving the conversation about you. Useful for spotting voices worth a closer look.

• SocialRead
  read_campaigns

  Lists your campaigns with their performance metrics, so marketing and support can compare what landed and what fell flat.

• SocialRead
  read_agent_metrics

  Returns per-agent replies, handle time and no-reply closes over a window of 1 to 90 days, 14 by default. Shows who handles what, and how fast.

• SocialRead
  search_knowledge

  Semantic search over your FAQs, policies and SOPs, the material replies draw on. Check what the knowledge base says before drafting an answer.

• SocialRead
  translate_text

  Translates a message into another language, with optional source language and tone, and returns the result without sending anything.

• SocialRead
  preview_autoreply_sweep

  Dry-runs the auto-reply sweep: shows which pending AI drafts would send under current autonomy and guardrails, and sends nothing. Test a change before you trust it.

• SocialWrite
  draft_reply

  Drafts an answer from your playbooks, knowledge, brand voice and guardrails, then returns it as a proposal. It never sends, so a person decides what goes out.

  Drafts for human review.
• SocialWrite
  assign_conversation

  Hands a conversation to a teammate, to AI, or back to the unassigned pool. Rebalances the queue without opening the inbox.

• SocialWrite
  set_conversation_status

  Moves a conversation between open, on hold, solved and closed, so queues reflect where each thread really stands.

• SocialWrite
  escalate_conversation

  Escalates a conversation to tier 2 or a lead, recording who escalated and why. Routes the hard cases upward with context attached.

• SocialWrite
  add_internal_note

  Posts a private note on a conversation for the team. The customer never sees it, so context travels with the thread.

• SocialWrite
  update_brand_voice

  Rewrites the voice replies follow: personality, things to avoid, always-rules, banned phrases, reading level and maximum length. It replaces the whole voice, so read the config first.

• SocialWrite
  upsert_playbook

  Creates or edits a WHEN-to-DO rule for replies, triggered by star range, sentiment, language and channel. Turns a judgment call into repeatable behavior.

  In auto mode this changes auto-reply behavior.
• SocialWrite
  upsert_guardrail

  Creates or edits a hard safety rule that every reply must respect, with an on-off switch. Guardrails bound what drafting and auto-reply are allowed to say.

• SocialWrite
  upsert_knowledge

  Adds or updates a knowledge source such as an FAQ, policy or SOP. The text is chunked and embedded automatically so replies can draw on it.

• SocialWrite
  upsert_macro

  Creates or edits a saved reply with a body, category, channels and tags, so common answers stay consistent across the team.

• SocialWrite
  upsert_known_issue

  Registers or updates a known issue with a status of active, monitoring or resolved. Replies about outages and bugs are grounded in these records.

• SocialWrite
  update_autonomy

  Sets the auto-reply dial from 0 to 100 percent, plus reply timing such as delay range and typing speed. Controls how much the AI may send without a person.

  Changes auto-reply behavior.
• SocialWrite
  propose_playbooks

  Scans recent conversations and drafts new suggested playbooks, saved as drafts for review. Turns recurring patterns in your inbox into rules you can approve.

  Saves drafts for human review.