Creating dynamic workflows for Airflow tasks present in a Python list

Question

I have a list of lists in the following way -

[['X_API', 'Y_API',....], ['Z_API', 'P_API', ...], [....], [...] .... ]

Here, each API name corresponds to a PythonOperator.

In Airflow, I would like to create task dependencies such that from a starting dummy task, I should have parallel tasks for each of the list inside the main list, and the operators inside the list of lists should execute in sequence :

How can I do this ? I would appreciate any help in this !

Existing code:

    args = {
            'depends_on_past': False,
            'start_date': datetime.now(),
            'email': '',
            'email_on_failure': False,
            'email_on_retry': False,
            'retries': 3,
            'retry_delay': timedelta(minutes=1)
        }   
    
    dag = DAG(dag_id, default_args=args, schedule_interval=None)
        
    with dag:
        tasks = []
        tmp, tmp2 = set(), set()
        Start = DummyOperator(
                task_id='Start',
                dag=dag
        )
    
        End = DummyOperator(
                task_id='End',
                dag=dag
        )
    
        for i in dags:
            for j in i:
                if 'APIs' in list(i.keys()):
                        
                    for l in i['APIs']:
                            tab = DummyOperator(
                            task_id=l['api'] + "_API",
                            dag=dag
                            )
                            tmp.add(tab)    
                elif 'tables' in list(i.keys()):
                        
                    for k in i['tables']:
                        tab2 = DummyOperator(
                            task_id=k['table'] + "_API",
                            dag=dag
                        )
                        tmp2.add(tab2)

      tasks.append(list(tmp))
      tasks.append(list(tmp2))

      for task in tasks:
        for op in range(0, len(task)-1):
            Start.set_downstream(task[op])
            task[op].set_downstream(task[op+1])
            task[op+1].set_downstream(End)