Today Pulsar's Autonomous Research Agent for Social and Media Intelligence becomes STORY, and moves from Private Beta into Public Beta. Over the past five months, Private Beta users pushed it from one-off questions into daily stakeholder briefings, overnight monitoring and reports that go straight to senior teams. You put it to work harder than we expected and that work shaped what we built, under a new name: STORY.
With STORY, you brief it once, the way you'd brief a researcher, and it stays on the job, watching, comparing, escalating what matters. STORY delivers finished work on its own clock, so you and your team can focus on interpretation, decision-making and strategy. It never clocks out.
So what's changed since the private beta?
Everything below is live today, shaped directly by your feedback.
AI for the judgement, not the bookkeeping
Most AI agent failures aren't reasoning failures. They're bookkeeping: date maths, recipient lists, filters, attribution and permissions. In this release, none of that is left to the model, so its judgement goes where it counts: what matters, and why.
Quotes are retrieved, not generated. STORY works on the raw posts in your search, not summaries of them. Every quote is the exact text of a post in your results, linked to its source and credited to the account that wrote it, not whoever reshared it. That includes replies, quote posts and reposts, where attribution usually breaks. No source post, no quote.
Dates are computed, not interpreted. Each schedule stores its window as a rule, such as "the 24 hours ending 07:30, London time". It works out the exact dates at run time, through time zones and clock changes. That covers rolling windows, "last N days" and calendar periods such as yesterday or last month. The model never does calendar maths.
Your settings aren't suggestions. Recipients, country, source and hour-by-hour filters apply exactly as set, on every run. Your search is resolved once, with no repeat lookups. The agent never decides who gets the brief.
It sees what you see. Nothing more. STORY runs with your Pulsar permissions, including private searches and folders shared with you, and can't reach beyond them.
Built to run without you
It runs on our servers, not in your session. Scheduled briefings run independently of your login. Close the laptop and the 7am brief still lands at 7am. If a step fails, it retries; the run doesn't start over.
You approve the plan once. Your goal, searches and timeframe are confirmed before anything runs. They're then carried into every turn and every scheduled run, so the fortieth turn still answers the question from the first.
It doesn't lose the plot. STORY now has a much larger working memory. Hours into a long analysis it still holds the brief, its sources and its early findings, so page twenty agrees with page two.
More control, better outputs
Quick filters in the composer: pick your search, date range and reply language without typing, or describe them in plain language, or mix the two.
Sharper verdicts: reports open with a verdict, a strength score from 0 to 100 and the change against the previous period, with up to four charts.
An easier-to-scan chat list: chats are titled by subject, a badge marks the ones with a schedule, and you can run several chats at once.
A fresh start: one thing to do by November 4th
These improvements apply to chats created from 28 September 2026 onwards. Chats and schedules set up before then were built on an earlier version of SAGA. They will no longer be supported, and these tend to cause most of the schedule failures some users have seen.
Please stop using them and create new ones, if needed. On November 4th all chats and schedules created before 28 September will be deleted.
To bring a chat or brief across:
Copy your brief. Open the old chat, copy your first prompt and any refinements, and save any reports you want to keep.
Start a new chat in STORY. Paste in your brief, set your search, date range and language with the filters, then confirm the plan.
Recreate any schedule. Set it up on the confirmed plan, add your recipients, then pause the old one so nobody receives duplicates.
Work only in the new chat. If something goes wrong in an older chat, start a new one. That's the fastest fix, and it lets our team spend their time improving STORY.
Chats created from 28 September onwards carry across automatically - you do not need to do anything.
If you run briefings for a wider team or senior stakeholders, message us in the app and we'll help you move them.
What's next
Public beta is where STORY keeps improving in the open. Coming up: a faster, more efficient analysis engine, a usage meter showing how much of your allowance remains, and better-formatted email briefings. We'll share updates as each one lands.
Getting access
If you want to try STORY, then speak to your account manager, or customer success manager and they can get you set up for a trial.
The updated STORY User Guide covers how to get the best results. Thank you to every Private Beta user who pushed STORY further. Keep the feedback coming, through in-app chat or support@pulsarplatform.com.
Getting started with STORY? These guides will help: