We were discussing how user activities are logged for insight by MongoDB. These activities include search terms, items viewed or wished, cart added or removed. orders submitted, sharing on social network and as impression or clickstream. Instead of using a time series database, the activity store and reporting is improved right from the box by MongoDB Moreover the user activities is used to compute user/product history, product map, user preferences, recommendations and trends. The challenges associated with recording user activities include the following
1) they are too voluminous and usually available only in logs
2) when they are made available elsewhere, they hamper performance.
3) If a warehouse is provided to record user activities, it helps reporting but becomes costly to scale
4) a compromise is available in NoSQL stores
5) but it is harder to provide a NoSQL store at the front-end where real-time queries are performed.
MongoDB addresses these with 
1) providing a dedicated store for large stream of data samples with variable schema and controlling the retention period
2) computing aggregations and derivatives separately
3) providing low latency to update data.
All user activity is recorded using HVDF API which is staged in User History store of MongoDB and pulled from external analytics such as Hadoop using MongoDB-Hadoop connector. Internal analytics for aggregation is also provided by Hadoop. Data store for Product Map, User preferences, Recommendations and Trends then store and make the aggregations available for personalization that the apps can use for interacting with the customer.
This completes the Data->Insight->Actions translation for user activities.


#codingexercise
We were implementing the furthest coprime of a given number such that it is in the range 2 – 250.
We made a single pass over the range of 2 to 250 to determine which ones were co-prime.
However we only need the farthest. Consequently we could split the range to be above the number or below the number so long as the input number is in the middle of the range and we could bail at the first encountered co-prime from either extremes. In the second iteration we can even compare the distance and exit when a coprime is located in the first iteration and the distance for that coprime is more than the current distance
int dist = 0;
// First Iteration as described yesterday
// Second Iteration as described yesterday
    // within second iteration
    if (dist > 0 && math.abs(n-i) < dist) {
        break;
    }