Organizing an academic conference or forum, the hardest part is often not the venue and schedule, but the list. Once the theme is set, the next step is to list guests—but where does the list come from? Most of the time it is cobbled together from several methods: a few people you can recall, a few you met at the last conference, a few recommended by colleagues. You manage to put one together, but you are not confident: Is this person still working in this direction? Have they changed institutions? Is there a more suitable person who has been missed?
This article is about replacing "cobbling together a list" with "verifying a list." MuchSci's Talent Profiles helps with the first half of this process: define directions by theme, compile publicly verifiable profiles one by one, verify recent ongoing projects and outputs, first form an evidence-based tiered list, and then hand it to the office assistant to create invitation materials.
01 How a list is assembled
Let's first look at a complete sequence. Its difference from "write down whoever comes to mind" is that it adds a verification step up front.
▲ Figure 1 | Identify people by theme, verify each one, then rank by tier
The first step is to define directions by topic. First write down the conference theme clearly, then break it into several subdirections—the theme is "new advances in such-and-such field," and in practice it should become individual small directions. "The more specific the directions, the more accurately you can identify people." Skip this step, and everything after it is based on gut feeling.
The second step is to define the candidate scope and first pull a batch of possible names for each direction. The third step is to compile a profile for each person, using four types of fields to examine research fields and current position; beyond the name, first confirm whether the institution and title match.
The fourth step is to verify recent ongoing research and look at the direction of projects led and outputs. This step is the easiest to skip: a name matching does not mean the direction still matches—some people worked in this direction years ago, but their current focus has shifted.
The fifth step is to form a tiered list, prioritize by fit and recent activity, and at the same time reserve the next in line. The sixth step is to produce invitation materials, handing the verified profile fields to the office assistant for drafting. Keep the sequence orderly, and every later step becomes easier.
02 Pulling together a list vs. verifying a list
The people on the two lists may be similar, but their reliability is different.
▲ Figure 2 | Listing names isn't hard; the hard part is being able to answer "why this person" for everyone on the list
When you directly pull together a list, the problems lie in several details. You cannot explain how the people were selected, mostly relying on impressions or a single recommendation from someone else; current positions are unverified, and the person may have already moved; it is unclear whether they are still working in this direction; when asked "why these people," you cannot give a clear reason; if someone cannot come, you can only improvise.
For a list verified one by one, each of these points has a counterpart. MuchSci profiles put public information into the same set of fields; the basis for selection is comparing research fields item by item by subdirection; current position is checked against the institution's public information using the employment history table; whether the person is still active recently is seen from the direction of projects led and outputs; for each person you can point to which column in the profile the basis falls under; after a rejection, the tiered list still has the next in line to fill in.
The list itself isn't long; what's long is the rationale behind each person. This is also why the verification step cannot be omitted: once the list is submitted, it represents the organizer's judgment and must hold up.
03 What three layers does a tiered list examine?
Tiering sounds like sorting, but it is actually a three-layer filter.
▲ Figure 3 | From top to bottom: Is the direction right? Are they still doing it? Can they come?
The first layer is topic fit, answering "Is the direction right?". Look at research fields, academic contributions, and projects led to judge whether the directions these people work in match this event's topics. The second layer is recent activity, answering "Are they still doing it?" Look at current institution, ongoing projects, and recent outputs to judge whether they are still active on the front line—this layer uses traces visible in public materials, does not represent a conclusion, and only narrows the possibilities to a smaller range.
The third layer is the part that requires manual completion: availability, travel, and willingness to attend. These three cannot be found in public materials, and profiles cannot help; they can only be asked in person. Distinguish the first two layers from the third so that you do not end up with "the list looks beautifully ranked, but when asked, none of them can come."
Here we should also explain the inclusion criteria of the profiles: they only include records related to national-level projects, and require the name, institution, and resume to accurately match public materials; mismatches are marked as "Pending Verification" and are not counted as experts. Applied to a guest list, the direction of this tradeoff is right—it is better to have one fewer person on the list than to have the awkwardness of "the materials don't match."
04 Functional interface: How profile fields feed into invitation materials
The interface is under the "Talent Profiles" tab of the agent assistant.
▲ Figure 4 | Basic information and education: a short profile plus a timeline; profile ID and data source are written at the very top
The top line is the public statement of this mechanism: profile ID, data source, and an explanation of the inclusion scope. The criteria are written in the profile header, meaning that when organizers are compiling the list, they can explain at any time how this material came about. Below are two sections: employment history and research outputs:
▲ Figure 5 | Employment history lists institutions and positions chronologically; research outputs are divided into academic contributions and projects led
These two sections are exactly the most commonly used parts of invitation materials. The institutions and positions in employment history are directly needed when writing the heading of the invitation letter; academic contributions and projects led in the main research outputs are the draft for introducing guests. The profile itself is organized as tables, so fields can go directly into materials without writing an introduction from scratch.
After verification, you can take away three things: a list that withstands follow-up questions, a ready-made draft of materials, and a ranked set of alternates.
Three things still need people to do: send invitations, confirm willingness, and arrange itineraries. Inviting is essentially communication, not retrieval—wording and timing are up to the organizer, only the person can answer whether they will attend, and availability and travel must be confirmed one by one. No matter how carefully you verify earlier, it cannot replace these later steps.
05 Boundaries
① This article discusses functional mechanisms and application methods and does not constitute a performance commitment. The completeness of profiles depends on the availability of public materials; the inclusion scope is limited and does not mean complete information can be provided for every researcher.
② Structured does not equal assessment. Profiles are the organization and presentation of public materials and do not constitute an evaluation of anyone's abilities, a recommendation, or an endorsement; whether to invite someone and in what capacity must be independently verified by the organizer, who bears responsibility.
③ Only public materials are used. No personal privacy data is involved; communication, agreements, and itinerary arrangements during the invitation process should follow the organizer's current rules and procedures.