Barry Clerjuste
AI Systems Engineer building open-source infrastructure for agentic AI, computational biology, and healthcare systems.
Mission
I build open-source AI systems for autonomous scientific discovery, computational biology, and healthcare infrastructure.
The work is focused on practical agentic systems: agents that reason over research literature, coordinate tools, plan experiments, support bounded clinical workflows, and produce outputs that humans can inspect, audit, reproduce, and improve.
Current research focus
raven-ai/
├── agentic-ai # tool routing, planning, multi-agent orchestration
├── computational-bio # omics, protein structure, literature-grounded workflows
├── healthcare-ai # consent, audit, evidence retrieval, clinical governance
├── local-first-ai # affordable deployment outside elite institutions
├── mobile-ai # on-device LLM inference via LiteRT-LM (Hermes Edge)
└── reproducibility # provenance, citations, tests, workflow traces
Raven ecosystem
| Project | Role | Status | Hugging Face |
|---|---|---|---|
| Raven AI | Flagship biology and healthcare agent platform | Active | Model · Space |
| OpenClinical AI | Healthcare deployment layer for Raven | Active | Model · Space |
| Home for AI | Local orchestration environment for AI workflows | Active | Model · Space |
| Hermes Edge | Mobile on-device AI agent (LiteRT-LM) | Active | Model · Space |
Technology stack
Languages Python · Rust · TypeScript · SQL · Bash
AI systems Agent orchestration · tool calling · retrieval · model routing
Bio + health FHIR · PHI-aware systems · omics workflows · evidence retrieval
Mobile AI LiteRT-LM · on-device inference · Apple ANE · Qualcomm QNN
Infrastructure Docker · Kubernetes · GitHub Actions · local-first deployment
Principles affordable · inspectable · evidence-linked · reproducible
Selected operating principles
Build systems that researchers can verify.
Prefer transparent workflows over black boxes.
Make scientific AI reproducible by default.
Treat healthcare AI as bounded support with audit trails, not magic.
Keep powerful tools affordable and deployable outside elite institutions.
Run AI where the data lives — on-device, sovereign, zero-trust.
GitHub activity
Collaboration
I am building the Raven ecosystem as an open-source research initiative for labs, researchers, startups, classrooms, clinical teams, and mobile users who need serious AI infrastructure without lock-in.
- GitHub: @simpliibarrii-crypto
- Hugging Face: bclermo
- Email: bclerjuste@gmail.com
- Site: https://simpliibarrii-crypto.github.io
Raven ecosystem: open-source infrastructure for agentic AI, computational biology, healthcare, and on-device intelligence.