A conference program holds a hundred talks or more, often in parallel sessions. You read the titles the night before, you miss the talk whose title did not say it was about your problem, and after the week you no longer know whom you meant to write to. The conference companion (Premium+) reads the whole program, keeps the talks close to your research, says why each one matters and what to ask, and keeps your program, your notes and your team together until the follow-up.

This guide follows a real program: Detection and Dynamics of Exoplanets (Coimbra, 7–11 July 2025), 120 talks and posters on Indico. We ranked it for a researcher who works on debris discs, the gas they hold and the planets that shape them. Every screenshot shows the result of that run, made on 10 November 2026.

1. Give the program and your research

Open the conference companion and paste the link of the program. An Indico event or a calendar feed (.ics, as Sched and Pretalx give) is read whole, with exact times, rooms and abstracts. Any other page, a PDF of the booklet or a pasted program is read by an AI.

Then describe your research in a few sentences. Separate your topics by a blank line: each one is looked for on its own, so a talk close to only one of them is not drowned by the others. You can also start from the profile of your daily paper alert, a project of your scoop watch or a collection of papers close to your work. Give your name and your co-authors' to set your own talks apart.

The form: the Indico link of the DDE 2025 program, and three topics of research separated by blank lines: debris discs as signposts of planets, gas in debris discs, the dynamical evolution of planetary systems.
The Indico link and three topics, each looked for apart.

Reading and ranking our program took under a minute.

2. Read your program day by day

The result says how many talks were kept: here 35 of the 120, after the AI read the 45 closest to our research. Each one has a level (Do not miss it, Worth it, If you have time), a sentence on why it matters for your research, and a question to ask its speaker. When a preprint of the talk is found, it is linked, with "Ask it questions" to chat with it before the session; the slides are linked when Indico has them.

The program of DDE 2025: 35 talks for you out of 120, the views of the program, the buttons to add it to a calendar, and the first talk of Monday 7 July with its level, why it matters and a question to ask.
35 talks kept out of 120; each with why it matters for you and a question to ask.

The reasons are written for the research you described, not as a summary of the abstract: a talk on spin–orbit measurements in a resonant chain is kept because a quiet or disrupted history "could also reshape planetesimal belts". Read them as a reason to look at the abstract, not as a judgement of the talk.

3. Mark the talks you go to, and take notes

Under each talk, mark Going, Seen or Not for me, and add a note, saved without leaving the page: it stays private. Once you mark talks as going, the calendar holds only those and your own; before that, it holds your program but the "if you have time".

A talk marked Do not miss it and Going, on the dynamical instability of multi-planet systems, with its preprint and a private note: ask about the fraction of planetesimals scattered inwards.
A talk not to miss, marked Going, with a note to prepare the question.

When two talks of your program are at the same time, the page says so on each and counts them at the top: choose, or ask a colleague to go. DDE 2025 had a single session, so our program has no clash.

By priority lists the same talks from the most important, whatever the day: the view to read the evening before. Whole program shows all 120 talks, to look for one the ranking left out.

The By priority view starts with a talk not to miss on breaking cold Jupiter resonance chains with stellar flybys, with its slides and its preprint.
By priority: the talks not to miss first.

4. Go with your team

Under Export, team and after, create a team link and send it to your colleagues. Those who join see the program and mark the talks they go to; on each talk you see who else goes, and the top of the program counts the talks not to miss that nobody from the team covers yet.

A talk not to miss on internal versus external perturbations, with Also going: Camille.
Camille, who joined with the team link, goes to this one.

A colleague can also rank the same program for their own research: the program is not read again, and the team still sees who goes where. Your notes and the research you described are never shown to them.

5. The people to meet

People to meet lists the presenters of the talks that matter most for you (20 at most), with their talks, their times and a link to their papers. Those who cite your papers come first, when you follow your papers in the citation alerts.

People to meet: Carolina Charalambous, Man Hoi Lee and Nader Haghighipour, each with their talk, its time and room, a link to their papers and a button to draft a follow-up email.
The presenters to meet, each with a follow-up email to draft after their talk.

After the talk, Draft a follow-up email writes a first message from the notes you took on their talks, to copy and edit; nothing is sent.

6. The themes of the conference

Themes groups all the talks by the closeness of their titles and abstracts, and shows how many of each are in your program. Here, 26 of the 35 talks kept are in Planetary System Dynamics, and only one in Exoplanet Detection Methods: the ranking followed the research we described, not the size of the sessions.

Five themes with bars: Planetary System Dynamics 45 talks, 26 for you; Exoplanet Characterization and Diversity 28, 3 for you; Exoplanet Detection Methods 23, 1 for you; Transit Timing Variations 19, 2; Young and Resonant Systems 5, 3.
Each theme with its talks, and those of your program.

7. Calendar, changes and follow-up

The Export section with the calendar subscription, the .ics file and the Markdown notes; the preprints to put in a collection; and Your team with its link and Camille in the team.
Export, preprints and team, in the last view of the program.

On your phone, the program reads the same way, and you mark talks and take notes during the sessions. During the conference, a Now view shows the talk going on, the next one and the rest of the day.

Two phone screens: the program by priority with its views, and a talk marked Going with its question and a note.
On a phone: the program by priority, and a talk marked Going with its note.

Limits

FAQ

Which programs can it read?
An Indico event or a calendar feed (.ics, as Sched and Pretalx give) is read whole, with the exact times and rooms. Any other page, a PDF or a pasted program is read by an AI, up to 400,000 characters (about 1,000 talks); the page says where the reading stopped.
Does the AI read every talk?
Every talk is compared with your research. The AI then reads the closest ones in full (at least 45, more for a large program) and says for each one why it matters to you, how much, and a question to ask.
Do my colleagues see what I wrote about my research?
No. A team link shows the program, the chosen talks and who goes where; a read-only share link shows the talks, why they were chosen and the people to meet, never who made the program nor the interests written.
What happens when the program changes?
An Indico program or a calendar feed is read again every day in the two weeks before the conference and every three hours while it goes on. When a talk of your program moves or is cancelled, or a talk as close to your research is added, you get an email, and the subscribed calendar follows.

Sources