This commit is contained in:
KS Jannette
2026-02-13 15:21:02 -05:00
parent 7f4c322280
commit 1a85c6e1e0

View File

@@ -1,16 +1,17 @@
# kongruity app # kongruity app
kongruity employs Large Language Model ("LLM") semantic grouping functionality to cluster large volumes of "to dos" or issue tags in development (or other) settings, according to thematic or topical similarity. kongruity employs Large Language Model ("LLM") semantic grouping to cluster large volumes of "to dos" or issue tags in development settings, by their thematic or topical similarity.
(To learn more about this topic, see, e.g., [Kozlowski A., Boutyline A., Semantic Structure in Large Language Model Embeddings Aug. 2025, arXiv:2508.10003v1:04 Aug 2025](https://arxiv.org/html/2508.10003v1)). (To learn more about semantic grouping, see, e.g., [Kozlowski A., Boutyline A., Semantic Structure in Large Language Model Embeddings Aug. 2025, arXiv:2508.10003v1:04 Aug 2025](https://arxiv.org/html/2508.10003v1)).
In kongruity world, "to dos", action items, agile tickets... the myriad artifacts of the creative/engineering process, are "sticky notes." In kongruity world, "to dos", action items, agile tickets, Jira comment threads (tagged for harvesting) ... the myriad artifacts of the creative-engineering process, are "sticky notes."
kongruity's React/Vite UI displays a board of seemingly chaotic "sticky notes". But with one click, they are transformed into manageable, actionable groups, each with a header that explains that group's semantic relation. kongruity's React/Vite UI displays a board of your team's seemingly chaotic "sticky notes". But with one click, they are transformed into manageable, actionable groups, each with a header that explains that group's semantic relation.
The backend is an Express API that serves "sticky note" data and proxies semantic grouping requests to Large Language Models. The backend is an Express API that serves "sticky note" data and proxies semantic grouping requests to Large Language Models.
Developers may freeely swap in other LLM SDKs and/or APIs... and alter prompt syntax at backend/services/clustering.service.js to compliement their experimentation with any LLM model/platform they prefer. Developers may freely swap in other LLM SDKs and/or APIs... and alter prompt syntax at backend/services/clustering.service.js to complement R&D with any LLM model/platform they prefer.
## Prerequisites ## Prerequisites