Understand code you didn’t write
Hud surfaces how all functions behave in production, directly in the IDE. Execution data, error rates, and performance trends on every function, with full execution path navigation.
Bring production context into your code
Every function in the codebase gets a Hudder: a compact overlay showing invocations per day, P90 duration, error rate and trends. Production telemetry, right next to the source code. No context switch, no need to leave the editor.
Trace any issue to its cause
The Hud panel displays the call graph based on actual production execution, not static analysis. See how execution time distributes across functions in a flow, identify failing paths, and follow execution paths from endpoint to root cause function.
Change code with confidence
Before editing a function, you see what’s at stake: how many flows depend on it, how often it runs, whether it’s stable or degrading. Know if a function is invoked 12 times a day or 5 million. Know whether a legacy module is actually executed or dead code. Know where exactly the time is spent across callers.
What changes for engineering teams
Try Hud in your IDE
FAQs
How do engineers typically access production data while coding?
Most workflows involve switching between the editor and separate tools: APM dashboards, error trackers, sometimes multiple browser tabs at once. Some IDE extensions surface basic error counts or deployment status, but function-level production behavior typically lives outside the editor. Runtime intelligence tools like Hud overlay runtime data directly on each function in the IDE, so engineers see production behavior without leaving their code.
Is there a way to see the runtime call graph of my code?
Most call graph tools rely on static analysis, tracing function calls by reading source code. These show what could be called based on code structure, but not what actually happens at runtime. A production-based call graph shows what is being called, how often, through which paths, and how execution time distributes across each function in the flow. Tools like Hud build the call graph from real execution data.
How can I identify dead code?
Static analysis and manual review can tell you what’s reachable in theory, but neither confirms whether a function actually runs in production. Code coverage tools help in testing environments, but test traffic doesn’t reflect real usage patterns. A newer category of tools, like runtime code sensors, surface production execution data directly in the IDE, making functions that are never invoked immediately visible alongside those that run millions of times a day, so refactoring decisions are based on actual usage rather than assumptions.
How does Hud help engineers write more production-aware code?
Most development happens without real-time knowledge of how the surrounding code behaves in production. Traditional workflows require checking dashboards or asking teammates for this context, if it’s checked at all. Hud surfaces execution frequency, duration trends, error rates, and caller relationships directly on each function in the editor, so engineers see what’s at stake before they make a change, not after it ships.
Trusted by engineers.
Human & artificial alike.
Hud runs on millions of services across massive production environments, with negligible overhead.