Digraph.dev VS AWAI

AWAI's answer:

Two things, and both come from how the analysis is built.

First, opinions and the reasons underneath them are gathered separately. Because of that, a reason shared by people who disagree still shows through, and a view only one person raised stays visible as an outlier instead of being averaged away.

Second, the same engine handles a group of one-on-one AI conversations and a live meeting between people. A meeting is sorted by topic while it is still running, and each participant reads it in the language they chose.

Both end in the same place: put a question to the material in ordinary prose and it becomes a report, exportable to PDF and editable by hand at no extra cost.

AWAI's answer:

Most tools in this space run AI interviews and hand back themes and sentiment. AWAI differs in three concrete ways.

It separates opinions from the reasons behind them, so you can see when people who disagree are working from the same reason — usually the thing that decides whether a decision holds.

It covers meetings as well as one-on-one interviews on the same engine, so work that starts as a survey and continues as a discussion stays in one place.

Reports are not a fixed deliverable. You put a plain-language question to the analyzed material and get a report angled for that reader — one for leadership, one for the floor, one as requirements for a vendor — up to 50 per project.

What AWAI does not have: a dedicated sentiment-analysis feature, and multimedia collection. Group analysis is text; audio is handled on the meeting side.

AWAI's answer:

HR and organization-development teams collecting employee opinions; executives and team leads who need decisions backed by the reasoning behind them; multilingual or distributed teams that meet across languages; consultants and researchers gathering qualitative feedback at scale.

AWAI's answer:

AWAI is a Japanese word — an old reading of the character for "interval" — meaning the space between two things. It names what the product looks at: the structure that shows up between one thought and another, and the common ground that appears between one person's thinking and another's.

The starting point was a shift that came with AI. Tools, techniques and knowledge — the things outside a person — became easy to produce, and once many people hold the same ones, they stop being what sets anyone apart. What AI can genuinely extend is the thinking on the inside.

Conversations are where that thinking lives, but they are hard to read afterward: things get said in the order they occur to people, not in the order that makes sense. AWAI takes conversations that have already happened and sorts them by what was said and why, so the shape of the thinking becomes something you can look at.

AWAI's answer:

Frontend: React, TypeScript, Vite, and Three.js (react-three-fiber) for the 3D views. Backend: Python, FastAPI, SQLAlchemy, and PostgreSQL with pgvector for embeddings. AI: Google Gemini — structured output for the analysis pipeline, embeddings for grouping, and the Live API for meetings. Real-time meetings: LiveKit. Auth: Firebase Auth. Infrastructure: Google Cloud Run and Cloudflare.