Dognosis

Date: 11-09-2025 HR Reached out in E-Mail also Yuvraj might have given referral

Round 1:

Casual 18mins talk Asked about past experience, asked about projects why you want to change your company ? i told engineering, he try to roast my previous company but i backed it, because i love scale.jobs how ambitious are you ? - I told pretty ambitious but not fast Asked tell me one topic that you will talk for hours ? - I told how nerds in previous generation are getting good now - I also told in my previous startup me and HR used to talk about caste system for hours, conversation got awkward here when i told caste system Asked how will you approach to learn something new ? He also asked projects around obsidian, there we found some mutual interest

Round 2:

Had a lot of internet issues of interviewer not me asked about me, asked about scale.jobs, asked about work i have done in scale.jobs asked about tell anything about you, I told, finance, geopolitics, engineering and math’s built an custom HashMap in JavaScript(he told to choose language of own) even though i wrote a blog, unable to remember a lot of things

// get(), set(), createHashmap()
// [{key: value}]
// hashmap -> createHashmap()(predefined indexes) ? -> get, set ?
// 100 -> 70 filled -> 130
function createHashmap(n) {
    const map = new Array(n);
    map.concat({ key: 1, value: 5 });
    return map;
}
 
function hashfunction(key, n) {
    return key % n;
} 
 
function get(map, key) {
    /*  for (let index = 0; index < map.length; index++) {
         const element = map[index];
         if (element.key == key) {
             return element.value;
         }
     } */
    const index = hashfunction(key, map.length);
    return map[index];
}
 
function set(map, key, value) {
    const index = hashfunction(key, map.length);
    console.log(index);
    map.splice(index, 0, value);
}
 
function main() {
    const map = createHashmap(10);
    set(map, 1, 5);
    console.log(map);
    console.log(get(map, 1));
}
 
main();

Later we discussed on collisions in HashMap he expected to write me the code in Golang, as he can ask questions of multi threading, but to my luck, i didn’t wrote in Golang and he stopped the implementation there asked him about thinking gaps, he released, I’m asking feedback, he told I’m good and have good knowledge of databases, he need time to re-evaluate and told all the best

Round 3:

He started telling about the tech stack that is being used in the company and the work i will mostly do in the company if i join
Python for backend that is being rewritten in rust now, frontend is in JavaScript that is slowly being migrated to Typescript, some components of frontend is written in php and in cloud they mostly rely on AWS and GCP. And If i join i will most probably will do is working with video processing and working on CCTV photo with rust and ffmpeg
Later he asked to write code for creating an queue in Golang

// [1,2,3,4,5]
// enqueue(value) --> [1,2,3,4,5] -> [nil,2,3,4,5] --> nil dequeue
// dequeue()
// [] -> [1] -> dequeue() -> read 1 or remove
// [] -> [2,1] -> dequeue()
package main
 
import "fmt"
 
func createQueue() []int {
    array := []int{}
    return array
}
 
func enqueue(arrayInstance []int, element int) []int {
    fmt.Println("Array Instance", arrayInstance)
    arrayInstance = append(arrayInstance, element)
    fmt.Println("Array Instance", arrayInstance)
    return arrayInstance
}
 
func dequeue(arrayInstance []int) []int {
    if len(arrayInstance) < 1 {
        return arrayInstance
    }
    arrayInstance = append(arrayInstance[:0], arrayInstance[:len(arrayInstance)-1]...)
    return arrayInstance
}
 
func main() {
    arrayInstance := createQueue()
    result := enqueue(arrayInstance, 1)
    fmt.Println(result)
    result2 := enqueue(result, 2)
    fmt.Println(result2)
    deleted := dequeue(result2)
    fmt.Println(deleted)
}

The things i got wrong while writing this code is

• I created multiple instances of the same array like

  func main() {
      arrayInstance := createQueue()
      result := enqueue(arrayInstance, 1)
      fmt.Println(result)
      result2 := enqueue(arrayInstance, 2)
      fmt.Println(result2)
      deleted:=dequeue(arrayInstance)
      fmt.Println(deleted)
  }

  I had noticed it and interviewer also hinted about it, thus made it right in the last
• He also wanted to instanciate the enqueue and dequeue methods using structs as methods, something like this

  package main
   
  import "fmt"
   
  // Queue struct
  type Queue struct {
  	data []int
  }
   
  // Create a new queue
  func NewQueue() *Queue {
  	return &Queue{data: []int{}}
  }
   
  // Enqueue method (adds element to the end)
  func (q *Queue) Enqueue(element int) {
  	q.data = append(q.data, element)
  }
   
  // Dequeue method (removes element from the front)
  func (q *Queue) Dequeue() (int, bool) {
  	if len(q.data) == 0 {
  		return 0, false // queue is empty
  	}
  	removed := q.data[0]
  	q.data = q.data[1:]
  	return removed, true
  }
   
  // Display method (optional helper)
  func (q *Queue) Display() {
  	fmt.Println(q.data)
  }
   
  func main() {
  	queue := NewQueue()
   
  	// enqueue elements
  	queue.Enqueue(1)
  	queue.Enqueue(2)
  	queue.Enqueue(3)
  	queue.Display() // [1 2 3]
   
  	// dequeue elements
  	if val, ok := queue.Dequeue(); ok {
  		fmt.Println("Dequeued:", val) // Dequeued: 1
  	}
  	queue.Display() // [2 3]
   
  	if val, ok := queue.Dequeue(); ok {
  		fmt.Println("Dequeued:", val) // Dequeued: 2
  	}
  	queue.Display() // [3]
  }

Follow up he asked to add concurrency to it, just to explain no need to implement
Later he asked how can i handle 2 threads to access the same resource

Later he asked about my projects about Snipzy and Semantic-Search-Engine
and in snipzy project he asked how can i extend that, i told i can add quality feature for the video, and he asked the problem with that, i told the loss of data in the video, he asked how can i solve that, i told i can use computer vision but mostly unsure about it, again he told, the loss of data is confirm, the thing is to minimize it that can be done by removing unwanted parts of the video and optimizing only the required parts of the video

In last he was happy with me