How to Reduce Python Code Complexity and runtime

The runtime of the code below is very huge. It takes 10 mins to process 6000 rows of data. Is there a way to reduce this huge runtime? My dataset has 200000 rows and it is not feasible to wait so long.

I tried executing it on Google Colab GPU and TPU as well. No luck. The runtime is still huge.

Please help me better understand alternatives to ifelse and the loops I used in this code and thereby reduce the runtime.

start_time= time.time()

folder= "April15"
curdir= source_dir+"//"+folder
files= sorted(os.listdir(curdir))

for file in files:
    print("extracting data from ", folder, file)
    if os.path.exists(dest_dir+r"//"+str(folder)+r"//"+file):
        print(file, ' exists')
        continue
        
    df = pd.read_csv(curdir +r"//"+ file, low_memory=False)
    n = len(df)
    print("jobs scrapped=", n)
    
    areas=[]
    titles = []
    companies = []
    ratings = []
    reviews = []
    experiences = []
    salaries = []
    locations = []
    descriptions = []
    naukri_categories = []
    posted_days = []
    tags = []

    df['footer']= df['footer'].replace('\nSave', '', regex=True)
    
    for i in range(n):
        
        head = df['header'][i].split('\n')
        if len(head)==5:
            title= head[0]
            company= head[1]
            rating= np.nan
            review= np.nan
            experience= head[2]
            salary= head[3]
            location= head[4]
        elif len(head)==7:
            title= head[0]
            company= head[1]
            rating= head[2]
            review= head[3]
            experience= head[4]
            salary= head[5]
            location= head[6]
        elif len(head)==6:
            title= head[0]
            company= head[1]
            rating= head[2]
            review= head[3]
            experience= np.nan
            salary= head[4]
            location= head[5]
        elif len(head)==4:
            title= head[0]
            company= head[1]
            rating= np.nan
            review= np.nan
            experience= np.nan
            salary= head[2]
            location= head[3]
        else:
            title= head[0]
            company= head[1]
            rating= np.nan
            review= np.nan
            experience= np.nan
            salary= head[2]
            location= np.nan
        
        job_descr = df['description'][i]

        foot= df['footer'].str.split('\n')
        if len(foot[i])==2:
            day= foot[i][1]
            category= foot[i][0]
        else:
            day= foot[i][0]
            category= np.nan

        areas.append(str(file[:-9]))
        titles.append(title)
        companies.append(company)
        ratings.append(rating)
        reviews.append(review)
        experiences.append(experience)
        salaries.append(salary)
        locations.append(location)
        descriptions.append(job_descr)
        posted_days.append(day)
        naukri_categories.append(category)

        if i%10000 == 0 and i != 0:
            print('\t', i, n)

    columns= ['functional area', 'title', 'company', 'experience','salary', 'location', 'description', 'naukri categories', 'posted_days', 'scraped_on', 'rating', 'review' ]
    df_new = pd.DataFrame(list(zip(areas, titles, companies, experiences, salaries, locations, descriptions, naukri_categories, posted_days, df['scraped_time'], ratings, reviews  )), columns =columns) 
    if len(tags)!=0:
        df_new['tags']= tags
    print('Extraction completed, writing to destination ', file)
    df_new.to_csv(dest_dir+r"//"+str(folder)+r"//"+file, index=False)
    
    print("time_taken for file=", (time.time()-start_time)/60, " minutes")
print("time_taken for folder=", (time.time()-start_time)/60, " minutes")strong text