AI Assistant Development for Optevo

Optevo is a modular collaboration platform that unifies communication, documents, tasks, and meetings. I led the development of its built-in AI assistant—designed to reduce context switching and give teams real-time insights directly inside their WorkPods.

Teams struggled with scattered communication threads, disorganized files, and time-consuming manual workflows. The assistant needed to understand user activity, summarize discussions, organize tasks, and surface insights while preserving strict privacy and role-based access control.

Designed a dual-layer system combining semantic understanding with secure, role-aware personalization.

Built contextual document + task understanding across all WorkPods.

Integrated RAG-based semantic search using Azure AI Search and OpenAI models.

Added inline writing assistance for summarization, paraphrasing, grammar correction, and clarity improvements.

Implemented smart nudges for deadlines, task creation, and content-quality suggestions.

Embedded directly into Optevo’s QuillJS editor:

Inline grammar + tone improvements

Predictive autocomplete

Meeting/discussion summarization

Non-intrusive coaching UI

Context-based task suggestions

Developed distributed AI microservices using Python (FastAPI + Flask).

Used Docker, PostgreSQL, and Redis for scalable indexing and real-time state.

Implemented JWT-secured, role-based access for private, personalized recommendations.

Built vector-based retrieval pipelines for user-specific file and message search.

The AI assistant transformed Optevo into a smarter collaboration environment—reducing manual work, surfacing insights instantly, and helping teams communicate and organize more effectively without leaving their workspace.