I think it would be nice to create a class representing the data on a single line, parse each line into an instance of your class, and then compare the objects¹.

Something like this:

List<Person> persons = new ArrayList<>();

for (String line : lines) { // Read the lines somehow

    String[] parts = line.split(", ");
    String position = parts[0];
    long id = Long.parseLong(parts[0]);
    // Et cetera

    persons.add(new Person(position, id, ...);
}

Then you can easily get all persons with tasks >= 3 in a for loop for example.

for (Person person : persons) {
    if (person.getTasks() >= 3) {
        // Print out the person
    }
}

By the way, a birthdate is best represented by a LocalDate. You can parse the date with

DateTimeFormatter formatter = DateTimeFormatter.ofPattern("dd,MM,yyyy");
LocalDate dob = LocalDate.parse(parts[3], formatter);

Grouping

Grouping is often done using a Map. You could map each employee position to a list containing the employees with that position:

List<Person> personsFromFile = ...;
Map<String, List<Person>> map = new HashMap<>();
for (Person person : personsFromFile) {

    // If the position does not yet exist as key in the map, create it
    if (!map.containsKey(person.getPosition())) {
        map.put(person.getPosition(), new ArrayList<>());
    }

    // Get the list with for this position and add the current person to it
    map.get(person.getPosision()).add(person);
}

Or using Java Streams API:

personsFromFile.stream()
   .collect(Collectors.groupingBy(p -> p.getPosision()));

¹ This is the whole point of object-oriented programming. We don't work with a bunch of variables, we model related data and functional classes and define functions operate on that object. Those are called methods. Each line in your file represents a person (or employee, you name it), so create a Person (or Employee) class.

Think in rows (objects) rather than columns

Java is object-oriented, so use objects. Rather than track each column of data in your data file, write a class to represent each kind of data found in a row. Each value in the row goes into a named member field on the class.

As your read each row, instantiate an object of that class, and populate its values. Add each new object to a List as you work your way brought the input file.

Tip: In real work, use a CSV library to help with reading and parsing such a comma-separated values file. You have a choice of such libraries in Java. Personally, I use Apache Commons CSV.

For the employee class

class employee {
    private String position;
    private long ID;
    private String Name;
    private String dob;
    private String title;
    private int task_done;

    public employee(String position, long ID, String Name, String dob, String title, int task_done) {
        this.position = position;
        this.ID = ID;
        this.Name = Name;
        this.dob = dob;
        this.title = title;
        this.task_done = task_done;
    }

    public long getID(){
        return ID;
    }

    public int getTask() {
        return task_done;
    }
}

Main

public static void main(String[] args) throws IOException {
    List<employee> employees = new ArrayList<employee>();
    BufferedReader inFile = new BufferedReader(new FileReader("Filepath.txt"));
    String inputline;
    while ((inputline = inFile.readLine()) != null) {
        String[] data = inputline.split(", ");
        employees.add(new employee(data[0], Long.parseLong(data[1]), data[2], data[3], data[4], Integer.parseInt(data[5])));
    }
    inFile.close();
    for (employee em : employees) {
        if (em.getTask() >= 3) {
            System.out.println(em.getID());
        }
    }
}

The file: (I have remove the commas for the Birthdate)

Manager, 987298347, Tesla, 03/04/1969, Mr, 4
Assistant, 290375020, Chris, 17/11/1989, Mr, 5
Manager, 99832482322, Steph, 11/02/1980, Ms, 4
Assistant, 679730283, Pete, 09/10/1980, Mr, 7

The output:

987298347
290375020
99832482322
679730283

You can modify it to do flexible tasks.