An AI company focused on compact models and workflow agents.
Tassili AI builds enterprise model systems around SLMs trained from scratch, domain-specific layers and controllable agents for production teams.
We believe the future of enterprise AI is not bigger models. It is precise ones.
Most enterprise work is recurring, bounded and domain-specific. It needs models that are small enough to run privately, specialised enough to be right, and governed enough to be trusted. That is what we build — with a small team, modest hardware and a lot of engineering.
Three enterprise building blocks.
SLMs
Compact models for repeated enterprise work that needs lower serving cost and better deployment fit.
DSLMs
Domain layers that increase precision for specific business functions and controlled contexts.
Agents
Workflow agents that retrieve, route, act and escalate under enterprise policies and review loops.
Organisations with real workflow volume and real governance needs.
We are not optimising for consumer usage. We are optimising for enterprise execution quality.
- Enterprise operations and shared services
- Support and customer operations teams
- Finance, compliance and documentation-heavy functions
- Companies that need private or hybrid deployment
From design partnership to production.
Discovery
Map workflow volume, domain constraints, data boundaries and automation opportunities.
Model system design
Choose the right model size, DSLM layer and agent boundaries.
Pilot
Validate output quality, execution reliability and operator adoption.
Production
Deploy with review loops, observability, governance and integration.
tassili-3b, 7b and 10b are ready.
See them run on your own workflows. We set up a private demo for your team — on our infrastructure or yours.