Novus is Pendo's product agent — an AI that connects directly to your codebase to automatically instrument, analyze, and improve your product with every release. It's built on Cortex, a living product context graph that keeps your intelligence grounded in what's actually running in production. For product managers, the core value is solving analytics drift — the silent problem where your dashboards drift away from your product as code ships faster than tracking can keep up. No manual tagging, no stale dashboards, no two-day analyst investigations.
What it does, how Cortex works, what it costs, and where it falls short.
Novus is the right tool for PMs on teams shipping at AI speed — where analytics drift is a real, ongoing problem and manual instrumentation can't keep up. Its codebase-connected auto-instrumentation is the feature no other analytics tool offers at this depth. Backed by Pendo's decade of product-analytics infrastructure, it has a strong foundation. It's in open beta, so expect rough edges. If you ship slowly and can maintain manual tracking, a traditional platform like Amplitude or a lighter tool like Vizr may suffice. If analytics drift is eating your data quality, Novus is purpose-built for you.
With AI, teams are shipping faster than ever. But the systems they use to understand what they shipped haven't changed — someone still has to manually tag features, maintain dashboards, and check them when they remember. When that someone falls behind, the data people use to make decisions quietly drifts away from the product actually running in production.
Pendo calls this analytics drift, and once you see it in a codebase, you can't unsee it. Routes get renamed. Features get refactored. New workflows ship. The instrumentation doesn't follow. Your dashboards look fine — they're just tracking a version of your product that no longer exists.
Novus is built for that gap. It connects to your codebase, understands the product through the code — no manual tagging required — and stays current as the codebase changes. At its core is Cortex, a living model of your product connecting code, behavior, and outcomes. It continuously observes, surfaces what matters without being asked, and connects insights to pull requests your team can review. Pendo has spent a decade building infrastructure to understand product behavior at scale — analytics, session replay, feedback, predictions — and Cortex is what keeps all of that grounded in reality as software changes.
Five features that separate a product agent from a product analytics dashboard:
Novus connects directly to your codebase and automatically instruments your product — understanding what it is, how it's structured, and how people use it. No manual tagging, no tracking plans, no feature dictionaries to maintain. When your codebase evolves, the instrumentation evolves with it.
At Novus's core is Cortex: a continuously updated model of your product that connects code, behavior, and outcomes in a single picture. It's what keeps the intelligence grounded in reality as software changes — and what connects Novus to Pendo's broader product intelligence layer built over a decade of analytics, session replay, and feedback infrastructure.
Analytics drift is what happens when your instrumentation rots as your codebase changes — routes get renamed, features get refactored, new workflows ship, and the tracking doesn't follow. Your dashboards quietly drift away from the product that's actually running in production. Novus is purpose-built to solve this: the instrumentation stays current with every release.
Novus surfaces what matters without waiting for someone to ask. Instead of a PM remembering to check a dashboard, the product agent continuously observes behavior and flags significant changes — so you find the activation drop the day it starts, not three weeks later when the cohort is already gone.
Novus connects insight to something reviewable and actionable — pull requests and in-product changes, not more reports. The team always decides what ships; Novus proposes, humans approve. This closes the gap between 'we found a problem' and 'we fixed it' without a separate ticketing round-trip.
Three concrete shifts in how product analytics work actually happens:
PM ships a new feature, but the analytics tracking was never updated. Three weeks later, the dashboard shows a mysterious drop — and nobody can tell whether it's a real problem or a broken event.
Novus auto-instruments the new feature as part of the release. The dashboard reflects what's actually running in production, and the 'mysterious drop' is either a real behavioral change (flagged with context) or doesn't happen at all.
PM asks a data analyst to investigate why activation dropped this week. The analyst spends two days querying the warehouse, finds the root cause, and writes a report that arrives after the cohort is already gone.
Novus continuously observes behavior and flags the activation drop the day it starts — with the root cause surfaced in plain language, not a two-day investigation cycle.
PM finds a friction point in the product and writes a Jira ticket. Engineering picks it up next sprint, ships the fix, and nobody circles back to measure whether it actually moved the metric.
Novus connects the insight to a pull request and measures whether the fix actually moved the metric — closing the loop from observation to action to outcome without a separate measurement step.
The caveat: Novus requires codebase access, which may be a constraint for teams with strict security policies. For the broader analytics-tool landscape, see our AI product analytics tools guide and our AI feature prioritization tools page.
AI product analytics that indexes every page, connects signals from support and feedback, and uses Discovery AI to find what to build. Broader signal coverage than Novus but lacks the codebase-connected auto-instrumentation. See our Productbet review.
AI-native product analytics with one script tag, no dashboards to build, and natural-language questions. Lighter and simpler than Novus — no codebase connection or auto-instrumentation, but faster to set up and cheaper. See our Vizr review.
A mature product analytics platform with AI-powered insights, session replay, and experimentation. More manual instrumentation than Novus but a deeper, battle-tested analytics suite. See our AI product analytics tools guide.
For the full analytics-tool landscape, see our AI product analytics tools guide and our AI product roadmap tools page.
50+ AI tools categorized by PM use case — prototyping, evaluation, research, analytics, roadmapping, and more. Each with a one-sentence description, a why-it-matters-for-PMs line, and a link. Free PDF, no credit card required.
Novus is in open beta and evolving rapidly — its Cortex capabilities, PR-connected workflows, and pricing will all shift in the coming months. This page captures it as of late 2026, but it can't tell you what shipped yesterday.
Signal Brief is the daily curation layer. Every weekday morning you get 5 new, vetted AI tools — each with a one-line description, a why-it-matters-for-PMs line, and a link. When Novus ships a major update or a new product-analytics tool launches, you'll read about it in your morning briefing, not months later.
A one-time deep dive on Novus. Read when you're evaluating auto-instrumentation and analytics-drift solutions.
5 new tools every weekday morning. 7-day free trial (25 tools), then $10/month or $96/year. The review is the map; Signal Brief keeps it updated.
Get 5 new AI tools for PMs every weekday morning. 25 tools during your free trial. Then $10/month or $96/year. Cancel anytime.
No credit card for the first 7 days. Cancel anytime.
Novus is Pendo's product agent — an AI that connects directly to your codebase to automatically instrument, analyze, and improve your product with every release. It's built on Cortex, a living product context graph that models what your product is, how it's structured, how people use it, and what that usage means for your business. Novus solves 'analytics drift' — the problem where your instrumentation rots as your codebase changes, leaving dashboards out of sync with what's actually running in production. It continuously observes behavior, surfaces what matters without being asked, and connects insights to reviewable pull requests. The team always decides what ships; Novus proposes, humans approve.
Analytics drift is what happens when your product ships faster than your instrumentation keeps up. Routes get renamed, features get refactored, new workflows ship — but the tracking plan doesn't follow. Your dashboards quietly drift away from the product that's actually running in production, so the data you're making decisions on is stale or broken without anyone realizing it. Novus solves this by connecting to your codebase and auto-instrumenting with every release, so the intelligence it generates always works from an accurate picture of what's actually live.
Pendo's existing platform provides dashboards, in-app guides, roadmaps, NPS, and session replay — a full product analytics suite that requires manual instrumentation and configuration. Novus is the AI agent layer on top: it auto-instruments by reading your codebase, continuously surfaces insights without you asking, and connects those insights to pull requests. Cortex — the product context graph at Novus's core — is what connects the agent to Pendo's broader infrastructure. In short, Pendo is the platform you configure; Novus is the agent that maintains it for you.
Productbet indexes every page, connects signals from support and feedback (Intercom, Slack, Zendesk), and uses Discovery AI to find what to build — broader signal coverage but no codebase-connected auto-instrumentation. Vizr is a lighter AI-native analytics tool with one script tag and natural-language questions — faster to set up and cheaper, but without the codebase connection or Cortex context graph. Novus's standout is auto-instrumentation via codebase connection: if analytics drift is your problem, Novus is the purpose-built answer. See our Productbet review and Vizr review for the full comparisons.
Novus is currently in open beta — check novus.ai or pendo.io for current access and pricing details. Pendo's full platform ranges from a free tier to enterprise plans, and Novus is positioned as the AI agent layer for teams shipping software with AI. For free alternatives, see our free AI tools for product managers guide.
Yes, but its full value unlocks for teams shipping fast. Novus's auto-instrumentation solves analytics drift regardless of whether you use AI coding agents — any team that ships frequently benefits from instrumentation that stays current. However, the pull-request-connected insight-to-action loop is most powerful when your engineering team ships at AI speed, because that's when analytics drift is most severe. If you ship slowly and can maintain manual instrumentation, a traditional analytics platform may suffice. If you ship fast and your tracking can't keep up, Novus is purpose-built for you.
It's in open beta, so expect rough edges, feature gaps, and evolving capabilities. Its codebase-connection model requires access to your repo, which may be a non-starter for teams with strict security or compliance constraints. And while auto-instrumentation is powerful, it can't replace the strategic judgment of deciding which metrics matter and what to do about them — Novus surfaces signals fast, but the PM still decides which ones deserve action. Treat it as infrastructure that keeps your data honest, not a replacement for product thinking.
5 new AI tools every weekday morning. 7-day free trial, then $10/month or $96/year. Cancel anytime. No sponsorships, ever.
Questions? Email [email protected]