mirror of
https://github.com/navanchauhan/ai-gong-data.git
synced 2026-04-03 07:38:44 -06:00
No description
| content_selectors | ||
| docs | ||
| .python-version | ||
| calls_extensive.py | ||
| generate_call_labels.py | ||
| generate_painpoint_tags.py | ||
| gpt_5_prompt.txt | ||
| import_sqlite.py | ||
| main.py | ||
| pyproject.toml | ||
| README.md | ||
| SCHEMA.md | ||
| users.py | ||
| uv.lock | ||
Gong Transcript Exporter
CLI helper around Gong's /v2/calls/transcript endpoint.
Quick Start
- Export credentials in your shell profile:
export GONG_ACCESS_KEY=... export GONG_SECRET_KEY=... export GONG_BASE_URL=https://api.gong.io - Grab yesterday's transcripts (local timezone) and write them to
dumps/:
The tool createsuv run python main.py --yesterday --prettydumps/transcripts-YYYY-MM-DD.jsonautomatically. - Include call metadata for the same range when needed:
This addsuv run python main.py --yesterday --include-calls --prettydumps/calls-YYYY-MM-DD.jsonalongside the transcripts. - Supply an explicit range when needed:
uv run python main.py --from 2025-11-01T00:00:00-05:00 --to 2025-11-02T23:59:59-05:00 \ --workspace-id 12345 --call-id 67890 --pretty --output custom.json
Load into SQLite
Normalize the JSON exports into SQLite for analysis:
uv run python import_sqlite.py \
--db dumps/gong.sqlite \
--calls dumps/calls-2025-ytd.json \
--transcripts dumps/transcripts-2025-ytd.json \
--clear-before
Schema:
callsstores metadata from/v2/calls(one row per call).call_transcriptsstores speaker/topic segments per call (positionpreserves order).transcript_sentencesstores each sentence with millisecond offsets linked to its segment.