Macroscope
AI-powered engineering intelligence platform that provides automated code review, commit status summaries, and an agentic assistant for engineering organizations.
At a Glance
About Macroscope
Macroscope is an AI-powered platform for engineering organizations that combines best-in-class code review, automated status reporting, and an agentic assistant that can investigate issues, write fixes, and ship PRs. Founded by Kayvon Beykpour, Joseph Bernstein, and Robert Bishop — the team behind Periscope (acquired by Twitter in 2015) and Magic Pony (acquired by Twitter in 2016) — Macroscope is backed by Lightspeed Venture Partners, Thrive Capital, and GV (Google Ventures).
What It Is
Macroscope is an engineering intelligence platform with three core pillars: Code Review, Status, and Agent. Code Review uses an agentic pipeline combined with purpose-built Abstract Syntax Tree (AST) walkers for common programming languages to catch real bugs before they ship. Status automatically generates succinct commit and PR summaries, sprint reports, and executive digests distributed via Slack. The Agent answers questions and takes action — available in Slack, GitHub, and via API — and can investigate failures, write code, and merge PRs autonomously.
How the Code Review Engine Works
Macroscope's code review combines an agentic pipeline with deterministic AST walkers built for the most common programming languages. This lets the agent gather rich context about every change without relying solely on an LLM to search for context, resulting in faster, more token-efficient analysis. The platform publishes an internal benchmark comparing its bug detection rate against tools like CodeRabbit, Cursor Bugbot, Greptile, and Graphite Diamond. According to Macroscope's own benchmark data, it achieves 48.31% bug detection (57/118 bugs) versus CodeRabbit at 45.76% and Cursor Bugbot at 42.37%. The platform also claims higher bug recall at a fraction of the cost compared to Claude's Code Reviewer, citing the Martian benchmark and official Claude pricing.
Key code review capabilities include:
- Correctness checks — finds real bugs without excessive false positives
- Auto-approve — automatically approves PRs with no bugs and minimal blast radius
- Fix It For Me — reply "fix it" on any comment; Macroscope branches, writes the fix, runs CI, and merges when checks pass
- Check Run Agents — define custom AI reviewers (with MCP access) that enforce stylistic conventions and workflows at PR review time
- CLI (beta) — run correctness review locally before pushing to a branch
- PR Summaries — every PR gets a clear, structured summary for reviewers
Status and Engineering Visibility
The Status pillar replaces status meetings with automated, on-demand engineering visibility. Macroscope generates succinct summaries of each commit and PR and distributes them to Slack in real time. Executive summaries are regenerated multiple times daily and organized by area (e.g., Auth & Security, Payments, Platform). Engineering productivity metrics show "pushed vs. landed" coding time, project focus breakdowns, and contributor-level stats — giving leadership visibility without interrupting engineers.
The Agent: Answers and Actions
Macroscope's Agent is available in Slack, GitHub, and via a webhook API. From Slack, teams can ask natural-language questions about product usage, activation funnels, or codebase changes and receive sourced answers (e.g., from PostHog). Macros enable recurring agentic workflows — such as daily standups, weekly bug reviews, or quarterly investor update drafts — triggered on a schedule with guardrails. The API allows teams to wire Macroscope into CI, observability, and support tooling, sending structured queries and receiving detailed investigation responses.
Integrations and Security
Macroscope offers native, first-party integrations with GitHub, Slack, Linear, Jira, Sentry, PostHog, Amplitude, and MCP servers. The platform is SOC 2 Type II compliant, encrypts all customer data in transit and at rest, architecturally isolates customer code, and does not train models on customer source code. Model provider agreements also prohibit training on customer IP.
Founding Team and Backing
Macroscope was founded by repeat entrepreneurs Kayvon Beykpour (co-founder of Periscope, former Twitter Consumer product and engineering lead), Joseph Bernstein (co-founder of Periscope), and Robert Bishop (co-founder of Magic Pony, an ML/computer vision startup). The company is backed by Lightspeed Venture Partners, Thrive Capital, GV (Google Ventures), and Adverb Ventures. The team describes itself as entrepreneurs and engineers who lived the engineering visibility problem firsthand and are building the product they wished they had.
Community Discussions
Be the first to start a conversation about Macroscope
Share your experience with Macroscope, ask questions, or help others learn from your insights.
Pricing
Teams
Usage-based plan with $100 free usage included for new workspaces. Includes Code Review, Status, and Agent.
- $100 free usage for new workspaces
- 1,000 free Agent credits per month
- Code Review: best-in-class bug detection, auto-fixes, auto-approvals, custom review rules
- Status: commit summaries, sprint reports, weekly digests, project classification, productivity stats
- Agent: answers and actions via Slack, GitHub, and API
Code Review
Usage-based code review at $0.05 per KB reviewed (10KB minimum per review).
- Best-in-class bug detection
- Auto-fixes issues
- Auto-approves safe PRs
- Custom review rules
- 10KB minimum per review
Status
Usage-based status reporting at $0.05 per commit.
- Commit summaries
- Sprint reports and weekly digests
- Project classification
- Productivity stats for devs and agents
Agent
Usage-based agent at $0.01 per credit. 1,000 credits/month included free.
- 1,000 credits/month included
- Writes code and ships PRs
- Available via Slack, GitHub, API
- Connects to your tools (PostHog, GCP, MCP)
Enterprise
Everything in Teams plus discounted pricing, priority support, custom agreements, and SOC 2 Type II report.
- Everything in Teams
- Discounted pricing for longer commitments
- Priority support
- Custom MSA & DPA agreements
- SOC 2 Type II Report
Capabilities
Key Features
- AI code review with bug detection
- Abstract Syntax Tree (AST) walkers for deep code context
- Auto-approve safe PRs
- Fix It For Me — automated fix, CI run, and merge
- Check Run Agents — custom AI reviewers per PR
- PR summaries
- CLI for local correctness review (beta)
- Commit and PR status summaries distributed via Slack
- Executive engineering summaries regenerated multiple times daily
- Sprint reports and weekly digests
- Project focus and productivity metrics
- Pushed vs. landed coding time tracking
- Slack-based agent for natural language Q&A
- Macros — scheduled agentic workflows
- Webhook API for CI and observability integration
- MCP server integration
- Per-repo enable/disable and file exclusion config
- Monthly spend limits and per-review caps
- SOC 2 Type II compliance
- Zero model training on customer code
Integrations
Demo Video

