Enriching your Bridge with data

Documentation Index

Fetch the complete documentation index at: https://hc.starbridge.ai/llms.txt

Use this file to discover all available pages before exploring further.

​

Overview

Bridges start as a list of items that your Search Agent has found. Adding columns enables you to enrich that base search with more information. The following outlines some tips on how to use add columns.

​

Use AI to Figure Out What Column to Use

Use AI to Generate Column Template option uses our AI to figure out what type of column you might want to select. This is done by adding a short statement to the prompt field along with any additional response examples you may have (for example the structure to an email). Depending on what your desired output is, you will be placed into the correct column type. Some example use-cases include:

• When is the due date of this RFP?
  • The Use AI column would pull out the Due Date attribute for the RFP and put it into its own column.
• Does the buyer anticipate funding shortages in the next year based on this meeting?
  • The Use AI column would pre-populate an AI Analysis column with a prompt to search the meeting signal for mentions of anticipated funding shortages in the next year.
• Does this buyer have any public mentions of challenges with AI?
  • The Use AI column would create a web agent that searches the internet for that buyer to find mentions of challenges with AI. https://www.loom.com/share/c3a77fb859a345c996af1c75dbc3986b?sid=9fe4cec1-ae80-410a-8135-60d53e9eec73

​

Column Types

​

Bridge Specific Columns

Bridge Specific Columns allow you to reference structured and standardized data that is specific to the Bridge Signal itself. See below for example attributes. You are able to select more than a single attribute at once, and each one will be displayed in its own column. These can be useful when used as inputs into AI-based prompts or for you to visually identify rows of interest.

​

Meetings Bridge Specific Columns

Most of these meetings bridge specific columns are simple text fields. Relevant Snippets is a special column type that extracts relevant quotes from the meeting document or transcript. Starbridge determines whether a quote is relevant by comparing it to your bridge description, search query, and organization context. When you or a consumer click the Relevant Snippets in the bridge results, email digest, or feed, you will see the relevant snippet and the video, audio, or text document at the timestamp or page where the relevant snippet can be found.

​

RFP Bridge and Purchase Bridge Specific Columns

​

Buyer Attributes Columns

Buyer Attribute Columns allow you to reference structured and standardized attributes that are specific to the Buyer itself (and free + collected by Starbridge). You are able to select more than a single attribute at once, and each one will be displayed in its own column. These can be useful when used as inputs into AI-based prompts or for you to visually identify rows of interest. Buyer attributes and enrichments are kept up to data automatically by Starbridge. For example, if a University’s enrollment changes and you have an enrollment column that syncs to your CRM, when Starbridge’s data team updates the enrollment of each institution, it will then update in your bridge and automatically sync to your CRM without intervention. For a list of all buyer attributes see here: All Buyer Data.

​

Contact Column

There are three types of contacts that can be added as a column to a Bridge. The first type is document based contacts. This option is only available for Meetings bridges and brings in the contact actually having the conversation in your Search. The second type is web based contacts. This option is available for all bridges. Web-based contacts give you the ability to describe the type of individual you are interested in finding contact information for. This may be useful if you know that your target contacts use different titles across different institutions or if the titles differ by State, or if you simply don’t know what the exact correct title is. You would describe the responsibilities of your target contact and then provide some examples or additional context that you think would be beneficial for Starbridge to know. Our AI will then check the Buyer’s website, any directories, any contact mentions across news articles, etc. to identify the best contact information. The goal of web-based contacts is to identify the individual who you want to speak with. While knowing who was leading the conversation you are interested in is valuable, that individual is not always the decision maker who determines whether or not to purchase your product. The third type is document based contacts with a web-fallback. This will return the contact identified in the Meeting, but we will run a web-based search to bring in any information not found in the Meeting. Just like document based contacts, document based contacts with a web-fallback is specific to Meeting bridges. For all contact types, you have the ability to display a specific field of the Contact Column in its own column if desired as a reference. Once the Contact Column returns its results, they will be displayed in a neat and formatted fashion. Clicking into each contact cell will give you a better sense of all of the contact information and the associated relevance/logic for why that particular contact was selected.

​

Summarized Relevance Columns

Summarized columns generate a summary of the signal and why it is relevant for you and your organization.

​

AI Analysis Column

The most popular use cases for AI analysis are to analyze items: IE: create a call script for a meeting discussion, create a score (score this set of accounts based on a set of criteria), generate a personalized email or summarize a set of items. The AI Analysis column type allows you to use natural language to describe what you are interested in learning about. AI can enrich the information that exists within the platform, allowing you to conduct analysis in bulk, quickly and effectively. While our web agents (see below) leverage the internet and web-based search, the AI analysis tool uses the context of the Meeting/RFP/Purchase and any other information in the table to instantaneously conduct analysis. Input Example: Output Example: You are able to reference other columns from your Bridge in your AI analysis. As with any AI-based prompting system, the more specific you are the better. Adding validation criteria (word limits, tone, etc.) and being highly specific in requesting the information you want to receive will help improve and refine the output from the AI analysis.

​

Web Agent Column

The web agent column is very similar to the AI Analysis column, however the web-agent searches for the information on the internet. It is not limited to the confines of the Starbridge platform, meaning it is able to identify pertinent and meaningful information with an almost unlimited scope. Customers use this analysis tool when they want to get supplementary information that cannot simply be found within the documents Starbridge has identified. Input Example: Output Example: As with any AI-based prompting system, the more specific you are the better. Adding validation criteria (word limits, tone, etc.) and being highly specific in requesting the information you want to receive will help improve and refine the output from the web agent.

​

Integration Column

Integration Columns allow you to sync specific columns from Starbridge to your CRM and reference specific data fields from your CRM in Starbridge. Visit the Syncing to Your CRM (SFDC or Hubspot) page to learn more.

​

Vendor Presence Column

If you want to know whether a given account has a relationship with a set of competitors, the competitor presence column is the best approach. The competitor presence column looks across our internal purchase orders, contracts, and web (optional) to determine whether there is evidence that a given buyer works with a competitor. Input the competitors and/or products you are interested in tracking, and we will output whether that buyer works with that competitor, any pricing detail, and an explanation of the relationship. Additionally, when creating a vendor presence column, you can choose how Starbridge handles gaps in the data:

• Use Only Confirmed Data — Show only vendor data explicitly found in Starbridge’s purchase order and contract records. This is the default setting.
• Use Confirmed & Inferred Data — Include all confirmed data and fill gaps with high-confidence estimates derived from purchase orders and contract signals. Inferred values are clearly flagged so you can distinguish them from confirmed data.

​

Edit Column (And Extracting Child Entities)

Referencing an existing column gives you the ability to extract a single attribute from a column containing multiple output fields in order to export the data in a more formatted way. Contact Columns are a perfect example of a column that contains multiple output fields that you may want to separate into their own columns.

​

Referencing Data from Another Bridge

Referencing an existing bridge enables you to enhance your data and knowledge of a list of Buyers. Using this cross bridge reference allows you to identify where and when a particular Buyer has appeared in other Bridges. You can cross-reference any bridges except for buyer bridges. One prominent use case for referencing data in another bridge is in account scoring when you might want to know:

• Do any of these buyers have RFPs matching specific search criteria currently posted that have not yet expired?
  • Reference an RFP Bridge
• Have these buyers been posting meetings where they express buying intent or pain points?
  • Reference a Meetings Bridge

When you add a reference bridge column, you will need to specify the following:

1. Which Bridge you want to cross-reference
2. Whether you want to reference new, actioned, not interested, or all mentions
3. What time-frame for referenced items do you want to include Some example use-cases include:
4. For my list of target cities in California (Buyer Bridge), I want to see all relevant Meetings from my Bridge which identified pain points (Meetings Bridge, all statuses, all dates).
5. For my list of target cities in California (Buyer Bridge), I want to see all actioned Purchases from my Bridge which identified contracts with competitors in the last 1 year (Purchases Bridge, actioned status, last year) The output of a existing bridge reference column is a column which identifies how many times (and where) a particular Buyer was referenced in your selected Bridge. In the example below, all the Buyers were not referenced in the selected Meetings Bridge, except for “City of Santa Paula”, which was referenced once:

Clicking into that “1” cell, you can see that City of Santa Paul appears 1 time in the selected reference Bridge. You can immediately view a summary of that meeting from the Buyer Bridge: If you click into “View Results in Reference Bridge”, you will be taken directly to the Meetings Bridge being used as the reference, with the relevant row (City of Santa Paula) automatically pre-filtered and viewable, along with all of the columns of data configured as part of the Meetings Bridge.

​

Testing & Running a Column

When you configure a new type of column, you can select how you release that data. Running for the first 5 rows may make sense if you want to test out your prompt and see if the results you’re getting back make sense and are in line with your expectations. Running all unrun rows will help fill in the gaps of any missing data you may have not run. And Run/Rerun All Rows will execute your functions/analysis on all of the rows! We recommend running the first 5 rows when you are testing out complex AI analysis or web-agent based prompts. Because the prompting process is naturally iterative, you should have the flexibility to see results, iterate, and rerun, until you get to the desired result/output, at which time you can run the remaining rows. Bridges Prompt Examples