Roadmap

Where we are headed

Five phases from foundation to scale — every milestone we've hit and everything we're building next.

Phase 01

Foundation

2023 Complete
Lab founded in Mogadishu
Initial Somali text collection (800K sentences)
First talent program cohort (20 engineers)
Research framework & open-source policy
Phase 02

Data & Evaluation

2025 Complete
SomNLP-Corpus v1 (1.5M sentences)
SomNLP-STT-Corpus (320 hours audio)
Somali Wikipedia Corpus (~11K articles)
Somali Names Dataset (92K names)
SomBench v0.1 evaluation suite (preview)
Phase 03

Models & Tools

2026 In progress
Goobo tokenizer (subword, Somali-optimised) — preview
whisper-som-small — ASR checkpoint (11.4% WER) — preview
som-en-mt — translation model (BLEU 24.3) — preview
SomNLP-Corpus v2 (887M+ tokens)
Soplang v2.0.0 — Somali-syntax programming language compiler
Goobo-base language model
Somali TTS voice (research preview)
Phase 04

Infrastructure

2026 – 2027 Planned
Public developer API for Somali models
RAG system for Somali knowledge retrieval
Online evaluation & model card platform
SomBench v1.0 (expanded, 12 tasks)
Community data-collection tooling
Phase 05

Scale & Expansion

2027+ Vision
100M+ sentence open corpus
Full instruction-tuned Somali LLM
Production-grade ASR (< 5% WER)
Regional expansion: Afar, Oromo, Tigrinya
Open API serving 1M+ requests/month
Community

Build it with us

Goobo Labs is a community effort. Whoever you are, there's a way to contribute.

Contribute on GitHub

Researchers

Use our datasets and benchmarks, co-author studies, and push low-resource NLP forward.

Developers

Build on open models and tokenizers, file issues, and ship Somali-language features.

Students

Join fellowships and bootcamps to learn ML and contribute to real research.

Contributors

Help collect, transcribe, and verify data — every contribution improves the models.

Build with us

Build Somali AI in the open

Use our datasets and models, contribute to the research, or partner with the lab. Everything we can open, we do.