AI Search Public Beta 1.2
Release Date: Friday, January 30, 2026
We’re excited to announce the release of Advanced Project Search (APS) — a major step forward in how Synthesis answers questions about projects and project-related relationships at your firm. This update will be automatically enabled for all current AI Search beta users at the close of business on Friday, January 30.
What’s Included in This Release
Better Answers for Project and Project-Related Relationship Queries
Advanced Project Search takes Synthesis AI Search to a new level. Synthesis AI Search will now search across your Employee, Project, Company, and Contact Directories. That means you can explore your connected data to answer complex, real-world questions like:
- What’s our average cost per square foot for higher education projects?
- Please list our LEED Silver or higher projects
- What projects have won AIA awards?
- Higher education projects in Massachusetts over 200,000 square feet.
- Projects with automated humidification controls
- Higher ed projects with LMNO consultants
- What architects have worked on projects in Washington?
- What higher ed projects have Em Davisson and Ann Johnson worked on together with BCN Engineering?
Because it connects across all these record types, Advanced Project Search can uncover relationships that were previously hidden—like which consultants and contractors you’ve collaborated with most often, which employees have worked together on specific project types, or which clients have the richest portfolios in your database.
Pro Tips
Pro Tip: Use the Project Directory as the Source of Truth for Structured Project Data
Advanced Project Search is optimized for structured project data coming from your Project Directory.
For the best results, project fields — such as project type, location, costs, dates, size, market, client, team members — should live in your Project Directory and flow from your connected systems (Deltek, Unanet, aec360, and OpenAsset).
While documents like past proposals still provide useful context for some Synthesis searches, Advanced Project Search will prioritize authoritative, repeatable project data sourced from the Project Directory when answering project-related questions.
This approach also makes data management more sustainable over time. When data needs to be corrected or improved, you know exactly where the source of truth lives and can improve it with confidence.
Troubleshooting
Troubleshooting: Missing Projects, Clients, or Consultants in Search Results
If certain projects, clients, or consultants don’t appear in Advanced Project Search results, check whether they are enabled to Show in Synthesis in your connected source system.
Only records marked Show in Synthesis are pulled into the Project, Client, and Contact directories. If a project, client, or consultant isn’t visible in those directories, it won’t be available to Advanced Project Search.
This is a common point of confusion during early adoption—especially for consultants or contacts—so it’s a good first thing to check if expected results are missing.
Troubleshooting: Missing or Unexpected Results Due to Unmapped Fields
Advanced Project Search can only reason over data that is mapped into Synthesis. If a field isn’t mapped, it won’t be available for project, company, or contact queries—and Synthesis won’t guess.
For example:
- Asking “Which higher education projects have we completed in the last five years?” requires a completion date field to be mapped.
- Asking “What lab projects have we done in Washington?” requires both a project type field and a location field to be mapped.
- Asking “Which structural engineers have we worked with in California?” requires contact or company role/type fields to be mapped so that Synthesis knows which companies are structural engineering firms.
If results seem incomplete or don’t make sense, it often means a relevant field hasn’t been mapped yet. As you test queries and gather user feedback, you may discover additional project, company, or contact fields that are worth bringing into Synthesis.
This is a normal part of early adoption and refinement—and mapping the right fields over time is what allows Advanced Project Search to get progressively smarter.
Troubleshooting: Numeric Fields Mapped as Text
Advanced Project Search can only perform calculations and comparisons on fields that are mapped as numbers.
If fields like square footage or construction cost are mapped into Synthesis as text instead of numeric or currency values, Advanced Project Search won’t be able to evaluate queries such as:
- Projects larger than 250,000 square feet
- Projects over $5 million in construction cost
To support these types of queries, size and cost fields need to be mapped using the appropriate numeric or currency data types.
If numeric-based queries return incomplete or unexpected results, this is a good place to check first.
Troubleshooting: Acronyms and Nicknames Not Returning Results
Advanced Project Search relies on the names and terms that exist in your data. If a client, company, or consultant is stored under its full legal name, searches using acronyms or nicknames may not return results.
For example:
- FEMA vs. Federal Emergency Management Agency
- VDOT vs. Virginia Department of Transportation
- AEI vs. Affiliated Engineers
If acronyms or commonly used shorthand don’t appear anywhere in Synthesis, Advanced Project Search won’t automatically infer the connection.
To address this, add acronyms and nicknames as synonyms in Search Optimization. This allows Advanced Project Search to recognize common variations and return more complete, intuitive results.
Troubleshooting: Acronyms or Nicknames for Project Types
If searches using shorthand or acronyms for project types don’t return expected results, the issue may be that those terms don’t exist in your structured data.
For example:
- Radiopharma vs. Radiopharmaceuticals
- NICU vs. Neonatal Intensive Care Unit
Advanced Project Search can only reason over the language it knows. If a nickname or acronym isn’t present in your data, it won’t automatically make the connection.
Adding these terms as synonyms in Search Optimization helps Advanced Project Search understand how your firm talks about its work and improves results for everyday, natural-language queries.
Troubleshooting: Duplicate or Overlapping Fields Can Create Confusion
Advanced Project Search works best when there is a single, clear source of truth for each type of project information.
We’ve seen cases where similar fields—such as Market Type and Market Type (Standard)—exist side by side, each populated differently. This can lead to confusing results, where the same project appears to belong to different categories depending on which field is being evaluated.
This kind of duplication can confuse Advanced Project Search, but it’s also confusing for employees who may not understand which field to rely on.
As a general guideline, aim for one field per domain of meaning. For example, all market-related information should live in a single, well-defined market field rather than being split across multiple overlapping ones.
Cleaning up and consolidating duplicate fields helps create clearer, more trustworthy results—and makes your project data easier to maintain over time.
Troubleshooting: Multiple Project Name Fields or Nicknames
Advanced Project Search uses a single, canonical Project Name field when reasoning about projects. At this time, it can’t evaluate multiple project name fields (such as long name, short name, alias, or nickname).
If your project database includes alternate project names that don’t live in the primary Project Name field, searches using those nicknames may not return expected results.
For now, the best way to support project nicknames or shorthand is to add them as synonyms in Search Optimization. This allows Advanced Project Search to recognize common variations without requiring changes to your underlying project database.
Support for additional project name fields may be explored in the future, but this release assumes a single, authoritative project name.
Troubleshooting: Project Data Stored Only in Documents
Advanced Project Search expects core project data to live in the Project Directory as structured fields.
If important project information exists only in documents—such as PDFs in a library—Advanced Project Search may not surface it reliably. For example, we’ve seen cases where project awards were stored only in proposal documents, making them difficult to find through project queries.
The same applies to other key project attributes, including specialized project types or classifications. If a data point is important for searching, filtering, or comparison, it needs to be captured as structured data on the project profile.
This follows a simple rule: Advanced Project Search looks to the Project Directory first. Documents provide helpful context, but structured project data belongs in the directory.
Help Us Keep Improving
Please encourage your users to continue rating their search results — it’s one of the most helpful ways to improve the quality of Advanced Project Search over time.
Additional Questions
Please email us at to support@knowledge-architecture.com.