> ## Documentation Index > Fetch the complete documentation index at: https://docs.aide.gg/llms.txt > Use this file to discover all available pages before exploring further. # Detection via DeltaMath > Understanding how teachers can detect AIde usage through DeltaMath's built-in analytics and how to mitigate these risks ## Primary Detection Method: Problem Completion Time ### The Main Risk The most significant detection risk comes from **teachers viewing your problem completion times** on their DeltaMath dashboard. This is the primary way a teacher might notice something unusual about your assignment completion. ### How It Works DeltaMath provides teachers with detailed analytics for each student's performance. When a teacher hovers over your individual attempts at solving a [problem](/concepts#3-problems-question-level), DeltaMath displays **the total time that was spent solving that specific problem**. This data is: * **Semi-difficult but also easy** for teachers to access (it's there if they look for it) * **Somewhat random** whether they'll actually check it * **Visible for every individual problem** you complete ### Risk Assessment by Problem Type The detection risk varies significantly based on the type of problems AIde is solving: #### Low Risk Problems (10-15 seconds) Problems that naturally take AIde's average of **10-15 seconds** to complete present **minimal detection risk**: * Reflecting points over an axis * Simple graphing tasks * Basic geometric transformations * Problems requiring visual analysis or drawing **Why they're safe:** These completion times appear natural for problems that genuinely require quick calculation. #### Higher Risk Problems Problems that should naturally take students longer but AIde completes at its default fast 10-15 second speed: * Complex algebraic equations * Multi-step calculations * Problems requiring multiple computational steps **Why they're riskier:** A teacher may notice that problems which typically require students to do complex calculations, move points around, or compute multiple steps are being completed suspiciously quickly. ### Impact Assessment This detection method is **not necessarily a massive issue** that prevents you from using the Free plan. However, it can become problematic if: * A teacher catches onto AIde and its capabilities * A teacher happens to notice your unusually short completion times * AIde's completing many complex problems quickly that should take longer ## Solution: Speed Regulator ### How It Works The **Speed Regulator** (available on Plus plan) helps you avoid detection by making your problem completion times look more natural. Here’s how it works: 1. **Arbitrary waiting:** AIde waits a random amount of time before submitting each answer. 2. **User-defined limits:** You set the minimum and maximum wait times. 3. **Random selection:** For each problem, AIde randomly picks a wait time within your chosen range. This creates realistic, varied completion time patterns that closely mimic typical student behavior. ### Configuration Example * Set minimum wait: 30 seconds * Set maximum wait: 180 seconds * AIde will randomly wait between 30-180 seconds per problem * Results in natural-looking completion patterns that match typical student behavior This completely eliminates DeltaMath's completion speed detection capabilities. ## Parallel Completion ### The Technical Reality There's one additional detection method that's **very advanced** and can only be identified by a teacher who is: * **Actively investigating** your account specifically * **Incredibly observant** with attention to timing details ### How This Detection Works If a teacher: 1. Navigates about **5 submenu clicks** deep into your account 2. Clicks on one specific assignment you completed 3. **Notes the completion time** of problems 4. Checks another assignment and compares timing 5. **They may notice** that problems within [sub-assignments](/concepts#2-sub-assignment-topic-level) are completed at almost exactly the same time ### Why This Happens On AIde's servers, we complete [sub-assignments](/concepts#2-sub-assignment-topic-level) **in parallel** (up to **5 simultaneously**) to speed up total assignment completion. This means: * Multiple [sub-assignments](/concepts#2-sub-assignment-topic-level) are processed at the same time * Problems within different [sub-assignments](/concepts#2-sub-assignment-topic-level) may show identical or very similar completion timestamps * The pattern becomes visible only through deep investigation ### Risk Assessment We are **comfortable with these limits** because: * It would be **incredibly difficult** for a teacher to notice via this method * Requires extensive manual investigation of your specific account * Most teachers don't have time for this level of detailed analysis * The detection pattern is subtle and easily missed ### Future Solution: Pro Plan We have a **Pro plan in development** that will allow adjusting the parallelism settings: * Control how many [sub-assignments](/concepts#2-sub-assignment-topic-level) are processed simultaneously * Add timing variation between parallel processes * Further reduce this already-minimal detection vector **Current Status:** This plan will supplement the Plus plan but is still in **very early development** and is **not a current priority**. ## Key Takeaways 1. **Primary risk** is completion time visibility - varies by problem type 2. **Speed regulator** completely eliminates timing-based detection 3. **Parallel processing** detection requires extensive teacher investigation 4. **Risk levels vary** significantly based on assignment content and teacher attention 5. **Most detection is avoidable** with proper plan selection and awareness Understanding these risks allows you to make informed decisions about when to use AIde and which plan best fits your detection avoidance needs.