NewBook a call
Cruq AI

AI agents for your business

Put AI agents to work across your business.

Production AI agents for sales, support, data, and voice, running on your own models and your own data. Ship in days, own everything you build.

Book a callStart buildingRead the docs →
live

Vertical AI products built on Cruq AI

Vertical AI products built on Cruq AI

talentreview.aifindprop.aiclawduck.com

Why Cruq AI

The agent platform you actually own, not another black box.

Own your stack

Bring your own provider keys and data. No markup, no lock-in, and nothing your agents touch ever leaves your boundary.

Ship in days

Compose agents for sales, support, data, or voice on a visual builder, then version and publish, from idea to production in days, not quarters.

Run in production

Ship agents behind one API with guardrails, tracing, and monitoring, so you always know what ran, what it cost, and why.

Cruq AI at a glance

100%

your keys and data stay in your workspace

1 API

call any published agent in production

0

vendor lock-in - own your agents and versions

AES-256

provider secrets encrypted at rest

SSE

responses stream token by token

RLS

row-level isolation per workspace

The platform

Build it, run it, and watch it - end to end.

Build

Design agents on a graph

Compose agent, tool, and router nodes on a canvas - system prompt, model, tools, and routing - then version and publish.

Assistant

openrouter/gpt-4o-mini

Tools

2 tools

end

Inference

Run on any model, your keys

Route agent inference through your own provider keys (OpenRouter, OpenAI, or any compatible endpoint). Secrets encrypted, scoped per workspace, guardrails applied.

providers /
openrouter/claude-3.5
openrouter/gpt-4o-mini
openai/gpt-4o

Observe

Trace every run

Capture every trace, input, and output in production. Filter, annotate, and replay any run, with cost and latency per model.

run_8f2a9c312mspass
run_3b7e4f891msfail
run_1c9d2e204mspass

Model library

Every model, one gateway.

Bring your own key and route any of these through your workspace. Swap models without touching your agents - no lock-in, no reselling.

OpenAI logoGPT-4oOpenAI
Anthropic logoClaude Sonnet 4.5Anthropic
Google logoGemini 2.5 ProGoogle
Meta logoLlama 3.3 70BMeta
Mistral logoMistral LargeMistral
DeepSeek logoDeepSeek V3DeepSeek
Qwen logoQwen2.5 72BQwen
xAI logoGrok 3xAI

...and any other OpenAI-compatible model, via OpenRouter, OpenAI, or your own endpoint.

Developer API

Run any agent with a single API call.

Publish an agent, mint an API key, and call it from anywhere. Responses stream back token by token over SSE - no SDK required.

Book a call ->
# run a published agent
curl https://cruq-87sjk09as.cruq.ai/v1/agents/AGENT_ID/runs \
  -H "Authorization: Bearer cruq_sk_..." \
  -H "Content-Type: application/json" \
  -d '{"input": "Draft a reply to this ticket"}'

event: RunContent
data: {"content": "Sure - here is a draft..."}

Built on Cruq AI

Production AI agents, shipped across industries.

The same builder, gateway, and monitoring we run at Cruq AI power a growing family of vertical AI products.

Recruitingtalentreview.aiAn agent-powered ATS that screens, ranks, and moves candidates for recruiters - plus an AI resume builder that helps applicants stand out.
Real Estatefindprop.aiAn agent-powered CRM that nurtures leads and closes deals for real estate agents - paired with a search engine that finds buyers the right property.
Assistantsclawduck.comAgent-powered personal assistants that handle your everyday tasks, schedule, and busywork - so you can focus on what matters.

Writing

From the lab

View all ->
goal + budgetGPUdatacenterNPUphoneMCU16 KBconvertoptimizequantizeprofile
August 19, 2026Research

Agentic model optimization for the edge

Getting a trained model to run fast on a phone, an NPU, or a microcontroller is a brutal, coupled search across the whole stack. The shift in 2025 and 2026 is handing that search to agents that work from a goal. A field guide to the four capabilities that make it work.

engineered trustreliabilityclarityexcellence
July 7, 2026Engineering

Quality Intelligence: engineering trust into AI features

Shipping AI features that hold up in production takes more than model testing. It takes a Quality Intelligence framework built on reliability, clarity, and measurable excellence.

prompt (instruction)vssourceactscoreupdateenvironment (practice)
May 12, 2026Research

Why RL environments beat prompt engineering for edge cases

Prompts are instructions. Environments are practice. Here's why the distinction matters when your agent keeps failing on the same class of inputs.

top 3 task types= 71% of token spendtask types by call volume
April 28, 2026Engineering

The hidden cost of frontier models in enterprise workflows

Most teams don't realize how much of their API bill comes from a small set of repetitive tasks. We traced the pattern across 12 deployments.

LLM call<1msring bufferin-memory, N=1000500msflushbatchedstoragerequest path returns immediately
April 14, 2026Infrastructure

Trace capture without slowing down your agent

Observability shouldn't be an afterthought. Here's our async capture architecture that adds less than 5ms overhead to any LLM call.

financeprecisionlegalconsistencyoperationscoverage
March 19, 2026Case Study

What we learned building private SLMs for three different verticals

Finance, legal, and ops all have different failure modes. Here's what we found when we trained domain-specific models for each.

Get started

Ready to build agents you own?

Put an agent to work for sales, support, data, or voice. Start free, or talk to us about rolling Cruq AI out across your team.

Book a call ->