[Expert Finder] Move experts from unstructured JSON toward a relational model

[nicktytarenko]

What changed

Relational experts (Expert + SearchExpert)

• Adds Expert: one row per normalized professional email, with registered_user, last_email_sent_at, and structured fields (name parts, academic title, affiliation, etc.).
• Adds SearchExpert: ordered membership linking an Expert to an ExpertSearch.

This mirrors behavior that existed in ARTEMIS, where a JSON blob was enough; in ResearchHub we normalize that data into proper tables so the same concepts are queryable, deduplicated, and consistent with the rest of the stack.

Why it matters

• Duplicate experts: easier to fix now that experts are keyed and stored in a structured table instead of only in per-search JSON.
• Exclusions: we stop using expert names for exclusion (fragile / ambiguous). excluded_expert_names is removed; excluded_search_ids (other search IDs) drives exclusion via resolved expert IDs from those searches.
• Invited experts: no separate invited-expert table. Expert.registered_user links a canonical expert row to a User when someone registers with the same email.
• Names from the LLM: the model used to return a single “full name” string, and suffixes like PhD could get glued onto the wrong part. Honorific, first, middle, last, suffix, and academic title are separate fields so we can build display/salutation strings (for now: first + middle + last as needed) without mis-parsing credentials.

Data migration

• Adds a management command to backfill Expert / SearchExpert from the legacy JSON field. It is temporary and can be removed after migration is complete and JSON is no longer the source of truth.

Important

Should be released together with ResearchHub/web#758

[codecov]

Codecov Report

❌ Patch coverage is 87.16356% with 62 lines in your changes missing coverage. Please review.
✅ Project coverage is 79.33%. Comparing base (4a83c39) to head (3596017).
⚠️ Report is 8 commits behind head on main.

Files with missing lines	Patch %	Lines
src/research_ai/serializers.py	74.24%	17 Missing ⚠️
src/research_ai/services/expert_persist.py	70.45%	13 Missing ⚠️
src/research_ai/services/expert_llm_table.py	92.74%	9 Missing ⚠️
src/research_ai/views/expert_finder_views.py	55.00%	9 Missing ⚠️
src/research_ai/services/expert_finder_service.py	82.50%	7 Missing ⚠️
src/research_ai/models.py	94.73%	2 Missing ⚠️
src/research_ai/services/expert_display.py	95.65%	2 Missing ⚠️
...rc/research_ai/services/email_generator_service.py	80.00%	1 Missing ⚠️
...c/research_ai/services/email_template_variables.py	80.00%	1 Missing ⚠️
...rc/research_ai/services/invited_experts_service.py	96.42%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3295      +/-   ##
==========================================
+ Coverage   79.28%   79.33%   +0.04%     
==========================================
  Files         619      623       +4     
  Lines       35224    35558     +334     
==========================================
+ Hits        27929    28211     +282     
- Misses       7295     7347      +52     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[chatgpt-codex-connector]

Codex usage limits have been reached for code reviews. Please check with the admins of this repo to increase the limits by adding credits.
Credits must be used to enable repository wide code reviews.

[sonarqubecloud]

Quality Gate passed

Issues
5 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud