Weave

• AI engineering impact measurement across development workflows
• Engineering output analysis based on complexity-weighted work
• AI usage and adoption tracking across coding tools
• Token intelligence for AI spending and consumption analysis
• AI-assisted code and pull request analysis
• Code quality scoring with Silk 1
• Weave Router for intelligent model selection and routing
• Wooly research agent for natural-language engineering analysis
• Engineering benchmarks combining AI, DORA, SPACE, and related metrics
• Integrations with source control, project management, communication, and AI coding tools

What Is Weave?

Weave is an engineering intelligence platform designed to help organizations understand software development from prompt to production. It combines large language models, machine learning, engineering telemetry, and development workflow data to measure output, quality, AI adoption, and cost in one environment.

The platform analyzes information such as commits, pull requests, reviews, deployments, AI activity, and token usage. This gives engineering leaders a broader view of how development work is progressing and how AI tools are contributing to the software delivery process. :contentReference[oaicite:4]{index=4}

Engineering Intelligence

Weave brings engineering activity into a unified measurement layer. Instead of relying only on metrics such as lines of code or pull request counts, the platform evaluates the nature and quality of engineering output alongside speed and AI involvement.

Its methodology includes complexity-weighted output measurements and connects AI spending with observed engineering work. Weave states that its output measurements are intended to provide context around cost, quality, adoption, and delivery rather than treating a single metric as a complete measure of AI return on investment. :contentReference[oaicite:5]{index=5}

AI Impact Insights

Weave provides visibility into how AI coding tools are being adopted and how their use relates to engineering output. Teams can examine AI-generated code, AI-assisted work, output, quality, and usage across different tools and engineers.

The platform tracks AI activity from tools such as Cursor, Claude Code, Devin, and other coding systems, allowing organizations to examine which tools are being used and how that activity relates to development outcomes. :contentReference[oaicite:6]{index=6}

Token Intelligence

Token Intelligence focuses on the cost and consumption side of AI-assisted development. Weave tracks token usage and spending so engineering organizations can understand where AI resources are being consumed and compare costs with measured output.

This can help teams examine AI spending by model, engineer, team, or workflow and connect usage data with broader engineering measurements. Weave’s research also uses production telemetry to analyze AI cost and output across engineering organizations. :contentReference[oaicite:7]{index=7}

Code Intelligence and Silk 1

Weave’s Code Intelligence uses its Silk 1 model to evaluate software development output. Silk 1 considers factors including intent, complexity, risk, and dependencies across files when estimating the amount of engineering work represented by a pull request.

The model provides a score along with plain-language reasoning and is designed to give engineering teams a consistent measurement scale for evaluating output. :contentReference[oaicite:8]{index=8}

Weave Router

Weave Router is an intelligent model-routing layer designed for coding agents. It evaluates incoming coding requests and routes them toward models based on factors such as task complexity, expected quality, and cost.

The Router can work with coding environments including Claude Code, Codex, and Cursor. Weave says teams can use existing provider subscriptions and route requests among supported models rather than manually selecting a model for every task. :contentReference[oaicite:9]{index=9}

The Router is available through a command-line installation and supports multiple model providers. Weave’s current pricing information states that Router usage is charged at 5% of routed spend, with teams of 50 or more able to work with a field engineering team for setup. :contentReference[oaicite:10]{index=10}

Wooly Research Agent

Wooly is Weave’s research agent for engineering analytics. Users can ask questions about their engineering organization in natural language rather than building individual dashboards or manually querying multiple data sources.

Wooly can analyze areas such as cycle time, AI-assisted output, engineering spend, revert rates, team performance, and AI tool usage. Its answers are grounded in connected organizational records and provide source links for the information used to produce an answer. :contentReference[oaicite:11]{index=11}

Natural Language Engineering Analysis

Wooly allows engineering teams to ask questions using everyday language. Questions can span different engineering datasets, including spending, output, review activity, and AI usage.

The research agent is available through the Weave workspace and can also be accessed through Slack, an editor environment, and the command line, allowing teams to bring engineering questions into existing workflows. :contentReference[oaicite:12]{index=12}

Engineering Benchmarks

Weave combines its AI measurements with established engineering metrics and frameworks. Its platform references DORA, SPACE, surveys, source-control activity, review data, deployment information, and AI telemetry to provide a broader view of engineering performance.

Weave also publishes research based on production telemetry from its customer base. Its Q2 2026 report analyzed data from 1,470 organizations and 21,409 engineers across AI impact, engineering output, legacy metrics, and prompt routing. :contentReference[oaicite:13]{index=13}

Integrations and Data Sources

Weave connects engineering information from source control, project management, communication, and AI development tools. Its published integrations include GitHub, GitLab, Bitbucket, Linear, Jira, Slack, and major AI coding tools such as Claude Code, Cursor, Codex, Copilot, and Devin. :contentReference[oaicite:14]{index=14}

By combining these sources, Weave can connect development activity with AI usage, spending, code quality, and delivery measurements rather than analyzing individual tools in isolation.

Security and Enterprise Controls

Weave states that its platform supports enterprise-oriented security and access controls, including SOC 2 Type II certification, GDPR compliance, HIPAA compliance, SSO through SAML and OIDC, and SCIM provisioning. :contentReference[oaicite:15]{index=15}

Who Can Use Weave?

Weave is designed primarily for engineering organizations that use AI coding tools and want to measure their impact on software development. Engineering leaders can use the platform to review output, AI adoption, quality, spending, and workflow performance, while engineering teams can use its analytics and research features to investigate specific development questions.

AI Cost and ROI Measurement

Weave connects AI spending with measured engineering output to provide a view of AI cost relative to work produced. Its methodology notes that this ratio does not by itself establish incremental revenue, hours saved, or a causal return on investment, making adoption, attribution confidence, and code quality additional considerations when interpreting the measurements. :contentReference[oaicite:16]{index=16}

Weave for AI Coding Workflows

For organizations using multiple AI coding assistants and models, Weave combines measurement and optimization features. AI Insights tracks adoption and output, Token Intelligence analyzes spending, Code Intelligence evaluates development work, Wooly provides natural-language analysis, and Weave Router can dynamically route coding requests between models.

Pricing and Availability

Weave provides a free way to get started, while some platform and enterprise offerings use paid plans or custom arrangements. The Weave Router currently lists pricing at 5% of routed spend. A specific starting price for the broader Weave engineering intelligence platform is not publicly disclosed. :contentReference[oaicite:17]{index=17}

Frequently Asked Questions

What is Weave used for?

Weave is used to measure engineering output, AI adoption, code quality, AI spending, and the effect of coding tools across software development workflows.

Does Weave measure AI coding activity?

Yes. Weave tracks AI coding activity and connects it with engineering output, code quality, spending, and other development measurements. :contentReference[oaicite:18]{index=18}

What is Weave Router?

Weave Router is a model-routing layer for coding agents that evaluates requests and directs them to suitable AI models based on factors including task complexity, cost, and expected quality. :contentReference[oaicite:19]{index=19}

What is Wooly in Weave?

Wooly is Weave’s research agent that answers natural-language questions about engineering organizations using connected engineering records and source-linked results. :contentReference[oaicite:20]{index=20}

Does Weave have a free option?

Yes. Weave provides a free starting option. Some services, including Weave Router, have usage-based pricing.

When was Weave launched?

Weave Engineering Intelligence was publicly launched on August 11, 2025. The company is listed as founded in 2024. :contentReference[oaicite:21]{index=21}

Engineering Work Gets Clearer Context

September 30, 2026

Review of Weave

Engineering teams can use Weave to connect AI usage with pull requests, code quality, output, and delivery metrics. Natural-language analysis adds context, though smaller teams may find its engineering analytics focus unnecessary.

Edward Collins

Share Your Experience: