Make project communication intelligent, not overwhelming
The problem is not a lack of communication. The problem is that important information is buried inside communication overload.
- Problem statement ID
- AS-02
- Organization
- ArchScale
- Category
- Web Application / Software / Automation / AI
- Theme
- Architecture, Interior Design & Construction — Communication Intelligence & Project Management
Background
Project communication is essential, but modern projects generate enormous amounts of it. Clients, architects, designers, contractors, vendors and project teams communicate through WhatsApp, emails, calls, meetings, voice notes, PDFs, screenshots and other channels.
The problem is not a lack of communication. The problem is that important information is buried inside communication overload.
A single conversation may contain a decision, several tasks, a deadline, an approval request and information relevant to multiple stakeholders.
Problem
Project teams often manually interpret conversations and convert them into tasks, follow-ups, meeting minutes and decisions.
Important information can therefore be:
There is a need for a system that can transform unstructured project communication into structured, actionable project information.
Expected solution
Participants should develop an intelligent communication layer capable of processing project conversations and extracting useful information.
Conversation Capture
Process text, transcripts, meeting notes, messages or other project communication.
Intelligent Summarization
Generate concise summaries of conversations or meetings.
Action Extraction
Identify tasks and actions from conversations.
Responsibility Detection
Identify who is expected to perform each action.
Deadline Detection
Extract deadlines or important dates.
Decision & Approval Extraction
Identify decisions, approvals and pending decisions.
Conversation-to-Task Conversion
Convert relevant communication into structured project tasks.
Searchable Project Memory
Allow users to find previous decisions, actions and discussions.
Illustrative example
“Rahul, please check the revised marble quotation with the client and get approval before Friday.”
The system could identify:
- Responsible
- Rahul
- Action
- Obtain client approval
- Subject
- Revised marble quotation
- Deadline
- Friday
- Type
- Approval
Key requirements
The prototype should demonstrate:
- Input from at least one communication source
- Processing of unstructured communication
- Extraction of useful project information
- Action / task identification
- Responsibility identification
- Summary or decision extraction
- Search / history or project memory
- A working end-to-end demonstration
AI may be used for summarization, extraction or classification.
Expected outcome
The goal is to reduce the amount of manual effort required to convert project communication into actionable information.