Readable before runnable
A workflow should make sense to the people responsible for the operation before software executes it.
What is SolveLang?
A readable, explainable workflow language for AI-assisted business processes.
SolveLang is an early-beta language and tooling project for making business rules, AI-assisted decisions, approvals, tools, ownership, and failure paths readable before they become managed automation.
Who it is for
Workflow intent first.
SolveLang is designed first for technical founders, hands-on operators, automation consultants, small agencies, and engineering teams that want workflow intent to remain understandable in source control.
It focuses on the layer that should be explicit before automation runs: what starts the process, which facts matter, where judgment changes the path, who owns the next action, what can fail, and when a person must review the result.
SolveLang is not another Zapier-style connector marketplace, a no-code mass-market automation builder, a production durable workflow engine, or a general-purpose autonomous AI workforce platform.
Version controllable
Workflow changes should be diffable, reviewable, attributable, and reversible like other source-controlled engineering work.
AI boundaries stay explicit
Deterministic rules, model-assisted judgment, tools, approvals, and failure paths should not be hidden behind one opaque automation step.
Safety is a product feature
Local validation, explicit capability boundaries, understandable failures, and human review matter before managed automation scales.
Current maturity
Working today, experimental, and planned are different things.
The Rust CLI is the canonical runtime. Workflow Intelligence Studio performs local-first deterministic analysis. The browser preview is intentionally smaller than the Rust runtime. Hosted API, billing, provider, and side-effect capabilities must not be interpreted as production readiness simply because code exists.
Working today
- Rust lexer, parser, AST runtime, interpreter, and CLI
- Run, validate, token, and AST inspection commands
- Source-located diagnostics and hardened execution modes
- Local-first deterministic Workflow Intelligence Studio analysis
- Browser-safe preview subset and public examples
Experimental / test-mode
- HTTP, file, and environment helpers
- AI-agent syntax and optional provider-backed responses
- Studio-to-SolveLang draft generation
- API keys, accounts, quotas, subscriptions, and billing infrastructure
Planned
- Stable language specification
- Production integrations and runtime adapters
- Full managed workflow execution
- Broader provider support
- Enterprise durability, governance, and observability
Use it today
Start with reproducible evidence.
Run the open-source Rust CLI, try the limited browser preview, inspect the public examples, or open Workflow Intelligence Studio for local deterministic analysis.
The guided workflow-audit service can also map one real process into decisions, missing branches, human review points, ownership, and an implementation-ready workflow specification.