Chesterfield Laboratories Inc
Kiln is an AI development workbench for teams building, evaluating, and optimizing AI systems. It connects evals, datasets, fine-tuning, RAG, agents, synthetic data, prompt optimization, and collaboration in a local-first workflow.
At a Glance
- Product teams building and iterating on AI systems
- Software and AI engineers
- Data scientists and ML practitioners
- Research teams
- +2 more
AI Tools by Chesterfield Laboratories Inc
(1)
Kiln
Desktop AI Development Workbench
Discussions
No discussions yet
Be the first to start a discussion about Chesterfield Laboratories Inc
Latest News
Products & Services
Local-first desktop workbench for building, evaluating, and optimizing AI systems, available for macOS, Windows, and Linux.
MIT-licensed open-source Python library used to load Kiln projects, run tasks, evaluate systems, build fine-tunes, work in notebooks, and integrate dataframes.
Opt-in server-side features including AI Assistant, auto-generated evals, Eval Builder, and Kiln Optimizer.
Visual tooling for document ingestion, search tools, retrieval-augmented generation, and synthetic data generation.
Market Position
Kiln positions itself as a local-first, Git-centric AI workbench spanning the full build-evaluate-optimize loop rather than a single-purpose orchestration, evaluation, or fine-tuning tool. Its differentiation is the combination of an inspectable desktop app, MIT-licensed Python library, privacy-preserving local data, collaborative Git workflows, and optimization capabilities; it competes for developer and LLMOps workflows with cloud-centric AI platforms and specialized eval, RAG, agent, and fine-tuning tools.
Leadership
Founders
Steve Cosman
Founder and CEO of Kiln; formerly worked on Apple Photos AI, was founder/CEO of ShoeboxApp, and worked at Microsoft. Kiln describes its founding team as engineers and scientists from Apple and Microsoft who shipped AI systems to billions of users.
Executive Team
Steve Cosman
Founder & CEO
Former Apple Photos AI engineer, founder/CEO of ShoeboxApp, and Microsoft engineer; his public profile describes him as building Kiln and previously working on AI systems at Apple and Microsoft.
Founding Story
Kiln was started after its team experienced the gap between mature software-engineering tooling and the needs of non-deterministic AI systems. The initial vision was to create the development platform they wished they had: a single, collaborative, observable workflow for building AI products with versioned data, regression-oriented evals, and rapid iteration.
Business Model
Revenue Model
Kiln keeps the core local-first app and open-source Python library free, and charges for Kiln Pro server-side capabilities such as AI Assistant, auto-generated evals, and optimization. Enterprise revenue comes from custom enterprise contracts covering SSO/SAML, procurement, SLAs, support, and onboarding.
Pricing Tiers
Open-source Python library, source-available desktop app, local datasets with Git sync, rate-limited Kiln Pro standard models, and community support.
Individual features plus enhanced models and higher limits, automatic agent optimization, priority access to new features, and email support.
Team features plus SSO/SAML, annual contracts, procurement support, SLA and priority support, dedicated solutions engineering, and custom onboarding/training.
Target Markets
- Product teams building and iterating on AI systems
- Software and AI engineers
- Data scientists and ML practitioners
- Research teams
- QA, product managers, and subject-matter experts
- Enterprise organizations requiring privacy, security review, SSO, and support
- Building and iterating on production AI applications
- Evaluating prompts, models, RAG pipelines, and agent behavior
- Fine-tuning and distilling smaller, cheaper models
- Generating synthetic training and evaluation data
- Building agents, tool-using systems, and MCP-enabled workflows
- Collaborative review by engineers, product managers, QA, and subject-matter experts
- Vector Institute for AI Research