Two tracks: live interaction and asynchronous analysis
In the current online flow, we are validating a general-purpose ASR track for detecting response boundaries alongside an asynchronous analysis track using our ASR and RAG.
Speech AI designed for children’s voices
We build technology that understands children’s voices better and can explain the evidence behind each result.
Internal validation7.05%
Child speech character error rate: 50 sentences from one held-out speaker, with no training overlap
Kaffol AI
From a child’s response to a report
We go beyond turning speech into text. The design considers age in months, the assessment items, and evaluation rules so that the evidence behind report statements can be reviewed.
- 01Child’s response
Child speech and behavior input
Children’s own responses captured through camera, microphone, and touch
- 02Speech recognition
Kaffol ASR+LoRA
Qwen3-ASR 1.7B base with our approximately 25 MB LoRA adapter
- 03Reviewing the evidence
Interpretation linked to evidence
Age-based items, peer comparison data, and evaluation rules reviewed by experts
- 04Explaining results
A clear, readable report
Describing current observations and next conversations based on actual responses
Internal validation results
Compared using the same child speech samples
We compared the character error rate (CER) of ASR with our own LoRA on the same internal samples. A lower value means a closer match to the original text.
Evaluation conditions50 sentences from one held-out speaker, with no training overlap
This small internal benchmark examines the effect of LoRA on children’s speech. It does not demonstrate superior performance across general use cases.
Character error rateLower is more accurate
| Model | Character error rate |
|---|---|
| Kaffol AI ASR | 7.05% |
| Qwen3-ASR 1.7B base | 8.89% |
| OpenAI gpt-transcribe | 9.56% |
How we describe our technology
We distinguish current work from ongoing research.
A small adapter for children’s speech
We are studying adaptation to children’s speech with an approximately 25 MB LoRA adapter that trains the text decoder while keeping the speech encoder fixed.
Exploring privacy through local AI
As field validation and manufacturing preparation progress, we are exploring local and on-device approaches that process children’s data closer to the device.
Standardization analysis and academic papers
Based on our item research, we are preparing statistical standardization analyses and academic papers. We do not describe unfinished research as complete.
Two patent applications relating to local AI
We have filed two patent applications relating to local AI. They are identified as applications, not granted patents.
Research with us