LoopGain Technologies Inc.
LoopGain provides open-source cost control for AI agent loops, using control theory to stop loops when they have converged and roll back when they degrade.
At a Glance
- AI Engineering Teams
- Agentic AI Developers
- Enterprise LLM Users
AI Tools by LoopGain Technologies Inc.
(1)
LoopGain
AI Agent Loop Cost Controller
Discussions
No discussions yet
Be the first to start a discussion about LoopGain Technologies Inc.
Latest News
LoopGain releases benchmark of 2,000 trials showing 92.8% reduction in API spend.
Hacker News discussion: Stop agent loops with control theory, not max_iterations.
Blog: How to design a strong verifier for AI agent loops.
LoopGain v0.1.0 initial release on GitHub.
Products & Services
An open-source (Apache-2.0) Python library for controlling AI agent loops using loop gain (Aβ) measurements.
A hosted telemetry receiver and dashboard for monitoring agent loop trajectories and performance.
A public benchmarking suite for measuring the cost and quality performance of AI agent loops.
Market Position
LoopGain positions itself as the 'Control' layer in agent tooling, contrasting with 'Observability' tools like LangSmith, Langfuse, and Helicone. It focuses on runtime intervention rather than post-hoc tracing.
Leadership
Founders
David Fitzsimmons
Senior Data & AI Product Builder with a focus on Applied AI Systems, Analytics, and iOS Apps. Based in Toronto, ON.
Executive Team
David Fitzsimmons
Founder
Senior Data & AI Product Builder based in Toronto, specializing in applied AI systems.
Founding Story
Started to solve the problem of 'runaway' AI agent loops that waste money by iterating past the point of improvement, using control engineering principles.
Business Model
Revenue Model
SaaS subscriptions for hosted telemetry and shared team workspaces.
Pricing Tiers
Full library, hosted dashboard for single-user, 7-day retention.
Shared workspace, per-iteration scrubber, share links, Read & ingest API, advanced alerts.
SSO, custom data residency, custom retention, 10M+/mo events, dedicated CSM.
Target Markets
- AI Engineering Teams
- Agentic AI Developers
- Enterprise LLM Users
- Verify-revise loops in agentic AI systems
- Cost control for LLM-based recursive tasks
- Preventing model degradation during iterative refinement
- Benchmarking agent loop efficiency