Spaces:
Running
Running
Test basic Streamlit app functionality
Browse files
app.py
CHANGED
@@ -1,399 +1,17 @@
|
|
1 |
import streamlit as st
|
2 |
-
import pandas as pd
|
3 |
-
import sqlite3
|
4 |
-
import os
|
5 |
-
from datetime import datetime
|
6 |
-
import time
|
7 |
-
from scraper import LinkedInScraper
|
8 |
-
from email_gen import EmailGenerator
|
9 |
|
10 |
-
# Configure Streamlit page
|
11 |
st.set_page_config(
|
12 |
page_title="Cold Email Outreach Assistant",
|
13 |
page_icon="📧",
|
14 |
layout="wide"
|
15 |
)
|
16 |
|
17 |
-
|
18 |
-
|
19 |
-
conn = sqlite3.connect('leads.db')
|
20 |
-
cursor = conn.cursor()
|
21 |
-
|
22 |
-
cursor.execute('''
|
23 |
-
CREATE TABLE IF NOT EXISTS scraped_data (
|
24 |
-
id INTEGER PRIMARY KEY AUTOINCREMENT,
|
25 |
-
linkedin_url TEXT UNIQUE,
|
26 |
-
company_name TEXT,
|
27 |
-
description TEXT,
|
28 |
-
industry TEXT,
|
29 |
-
website TEXT,
|
30 |
-
scraped_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
|
31 |
-
)
|
32 |
-
''')
|
33 |
-
|
34 |
-
cursor.execute('''
|
35 |
-
CREATE TABLE IF NOT EXISTS generated_emails (
|
36 |
-
id INTEGER PRIMARY KEY AUTOINCREMENT,
|
37 |
-
linkedin_url TEXT,
|
38 |
-
recipient_name TEXT,
|
39 |
-
recipient_email TEXT,
|
40 |
-
company_name TEXT,
|
41 |
-
subject_line TEXT,
|
42 |
-
email_body TEXT,
|
43 |
-
tone TEXT,
|
44 |
-
creativity REAL,
|
45 |
-
quality_score REAL,
|
46 |
-
generated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
|
47 |
-
FOREIGN KEY (linkedin_url) REFERENCES scraped_data (linkedin_url)
|
48 |
-
)
|
49 |
-
''')
|
50 |
-
|
51 |
-
conn.commit()
|
52 |
-
conn.close()
|
53 |
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
return EmailGenerator()
|
59 |
-
except Exception as e:
|
60 |
-
st.error(f"❌ Failed to load AI model: {str(e)}")
|
61 |
-
st.info("💡 The model will be downloaded automatically on first run. Please ensure you have a stable internet connection.")
|
62 |
-
return None
|
63 |
|
64 |
-
|
65 |
-
"""Process CSV data and generate emails"""
|
66 |
-
scraper = LinkedInScraper()
|
67 |
-
email_gen = load_email_generator()
|
68 |
-
|
69 |
-
if email_gen is None:
|
70 |
-
return None
|
71 |
-
|
72 |
-
results = []
|
73 |
-
|
74 |
-
# Progress tracking
|
75 |
-
progress_bar = st.progress(0)
|
76 |
-
status_text = st.empty()
|
77 |
-
|
78 |
-
for idx, row in df.iterrows():
|
79 |
-
progress = (idx + 1) / len(df)
|
80 |
-
progress_bar.progress(progress)
|
81 |
-
status_text.text(f"Processing {idx + 1}/{len(df)}: {row['name']} at {row['company']}")
|
82 |
-
|
83 |
-
# Scrape company data
|
84 |
-
with st.spinner(f"🔍 Scraping data for {row['company']}..."):
|
85 |
-
company_data = scraper.scrape_company_data(row['linkedin_url'])
|
86 |
-
|
87 |
-
# Generate emails with multiple variations
|
88 |
-
variations = []
|
89 |
-
for i in range(num_variations):
|
90 |
-
with st.spinner(f"✍️ Generating email variation {i+1}/{num_variations}..."):
|
91 |
-
email_data = email_gen.generate_email(
|
92 |
-
company_data=company_data,
|
93 |
-
recipient_name=row['name'],
|
94 |
-
recipient_email=row['email'],
|
95 |
-
tone=tone,
|
96 |
-
creativity=creativity
|
97 |
-
)
|
98 |
-
if email_data:
|
99 |
-
variations.append(email_data)
|
100 |
-
|
101 |
-
# Combine all data
|
102 |
-
result = {
|
103 |
-
'name': row['name'],
|
104 |
-
'email': row['email'],
|
105 |
-
'company': row['company'],
|
106 |
-
'linkedin_url': row['linkedin_url'],
|
107 |
-
'company_description': company_data.get('description', 'N/A'),
|
108 |
-
'industry': company_data.get('industry', 'N/A'),
|
109 |
-
'website': company_data.get('website', 'N/A'),
|
110 |
-
'variations': variations,
|
111 |
-
'best_variation': max(variations, key=lambda x: x.get('quality_score', 0)) if variations else None
|
112 |
-
}
|
113 |
-
results.append(result)
|
114 |
-
|
115 |
-
# Small delay to be respectful to servers
|
116 |
-
time.sleep(1)
|
117 |
-
|
118 |
-
progress_bar.progress(1.0)
|
119 |
-
status_text.text("✅ Processing complete!")
|
120 |
-
|
121 |
-
return results
|
122 |
-
|
123 |
-
def display_results(results, tone, creativity):
|
124 |
-
"""Display results in an interactive table format"""
|
125 |
-
if not results:
|
126 |
-
st.warning("No results to display.")
|
127 |
-
return
|
128 |
-
|
129 |
-
st.subheader("📊 Results Overview")
|
130 |
-
|
131 |
-
# Summary metrics
|
132 |
-
col1, col2, col3, col4 = st.columns(4)
|
133 |
-
|
134 |
-
total_leads = len(results)
|
135 |
-
successful_generations = sum(1 for r in results if r['best_variation'])
|
136 |
-
avg_quality = sum(r['best_variation']['quality_score'] for r in results if r['best_variation']) / max(successful_generations, 1)
|
137 |
-
|
138 |
-
col1.metric("Total Leads", total_leads)
|
139 |
-
col2.metric("Successful Generations", successful_generations)
|
140 |
-
col3.metric("Average Quality Score", f"{avg_quality:.1f}/10")
|
141 |
-
col4.metric("Success Rate", f"{(successful_generations/total_leads)*100:.1f}%")
|
142 |
-
|
143 |
-
# Results table
|
144 |
-
st.subheader("📋 Generated Emails")
|
145 |
-
|
146 |
-
for idx, result in enumerate(results):
|
147 |
-
with st.expander(f"📧 {result['name']} - {result['company']}", expanded=False):
|
148 |
-
if not result['best_variation']:
|
149 |
-
st.error("❌ Failed to generate email for this contact")
|
150 |
-
continue
|
151 |
-
|
152 |
-
best = result['best_variation']
|
153 |
-
|
154 |
-
# Three columns layout
|
155 |
-
col1, col2, col3 = st.columns([1, 2, 1])
|
156 |
-
|
157 |
-
with col1:
|
158 |
-
st.write("**Contact Info:**")
|
159 |
-
st.write(f"👤 **Name:** {result['name']}")
|
160 |
-
st.write(f"📧 **Email:** {result['email']}")
|
161 |
-
st.write(f"🏢 **Company:** {result['company']}")
|
162 |
-
st.write(f"🏭 **Industry:** {result['industry']}")
|
163 |
-
if result['website'] != 'N/A':
|
164 |
-
st.write(f"🌐 **Website:** {result['website']}")
|
165 |
-
|
166 |
-
# Quality indicator
|
167 |
-
quality = best['quality_score']
|
168 |
-
if quality >= 8:
|
169 |
-
st.success(f"🌟 Quality: {quality:.1f}/10")
|
170 |
-
elif quality >= 6:
|
171 |
-
st.warning(f"⚡ Quality: {quality:.1f}/10")
|
172 |
-
else:
|
173 |
-
st.error(f"⚠️ Quality: {quality:.1f}/10")
|
174 |
-
|
175 |
-
with col2:
|
176 |
-
st.write("**📧 Generated Email:**")
|
177 |
-
|
178 |
-
# Subject line
|
179 |
-
st.write("**Subject:**")
|
180 |
-
st.code(best['subject_line'], language=None)
|
181 |
-
|
182 |
-
# Email body
|
183 |
-
st.write("**Email Body:**")
|
184 |
-
st.text_area(
|
185 |
-
"Email Content",
|
186 |
-
best['email_body'],
|
187 |
-
height=300,
|
188 |
-
key=f"email_body_{idx}",
|
189 |
-
label_visibility="collapsed"
|
190 |
-
)
|
191 |
-
|
192 |
-
with col3:
|
193 |
-
st.write("**🎯 Generation Settings:**")
|
194 |
-
st.write(f"**Tone:** {tone}")
|
195 |
-
st.write(f"**Creativity:** {creativity}")
|
196 |
-
|
197 |
-
st.write("**📊 Company Data:**")
|
198 |
-
if result['company_description'] != 'N/A':
|
199 |
-
with st.expander("📄 Description", expanded=False):
|
200 |
-
st.write(result['company_description'][:200] + "..." if len(result['company_description']) > 200 else result['company_description'])
|
201 |
-
|
202 |
-
# Show all variations if multiple
|
203 |
-
if len(result['variations']) > 1:
|
204 |
-
st.write(f"**🔄 Variations ({len(result['variations'])}):**")
|
205 |
-
for i, var in enumerate(result['variations']):
|
206 |
-
quality_color = "🌟" if var['quality_score'] >= 8 else "⚡" if var['quality_score'] >= 6 else "⚠️"
|
207 |
-
st.write(f"{quality_color} Variation {i+1}: {var['quality_score']:.1f}/10")
|
208 |
-
|
209 |
-
def export_to_csv(results):
|
210 |
-
"""Export results to CSV format"""
|
211 |
-
if not results:
|
212 |
-
return None
|
213 |
-
|
214 |
-
export_data = []
|
215 |
-
for result in results:
|
216 |
-
if result['best_variation']:
|
217 |
-
best = result['best_variation']
|
218 |
-
export_data.append({
|
219 |
-
'name': result['name'],
|
220 |
-
'email': result['email'],
|
221 |
-
'company': result['company'],
|
222 |
-
'industry': result['industry'],
|
223 |
-
'website': result['website'],
|
224 |
-
'subject_line': best['subject_line'],
|
225 |
-
'email_body': best['email_body'],
|
226 |
-
'quality_score': best['quality_score'],
|
227 |
-
'linkedin_url': result['linkedin_url'],
|
228 |
-
'company_description': result['company_description']
|
229 |
-
})
|
230 |
-
|
231 |
-
return pd.DataFrame(export_data)
|
232 |
-
|
233 |
-
def main():
|
234 |
-
# Initialize database
|
235 |
-
init_database()
|
236 |
-
|
237 |
-
# Header
|
238 |
-
st.title("📧 Cold Email Outreach Assistant")
|
239 |
-
st.markdown("Transform your lead list into personalized, high-converting cold emails using AI")
|
240 |
-
|
241 |
-
# Sidebar for settings
|
242 |
-
with st.sidebar:
|
243 |
-
st.header("🎛️ Settings")
|
244 |
-
|
245 |
-
# Email generation settings
|
246 |
-
st.subheader("📝 Email Generation")
|
247 |
-
tone = st.selectbox(
|
248 |
-
"🎭 Tone",
|
249 |
-
["Professional", "Friendly", "Direct", "Casual", "Formal"],
|
250 |
-
index=0,
|
251 |
-
help="Choose the tone for your emails"
|
252 |
-
)
|
253 |
-
|
254 |
-
creativity = st.slider(
|
255 |
-
"🎨 Creativity Level",
|
256 |
-
min_value=0.1,
|
257 |
-
max_value=1.0,
|
258 |
-
value=0.7,
|
259 |
-
step=0.1,
|
260 |
-
help="Higher values = more creative but potentially less focused emails"
|
261 |
-
)
|
262 |
-
|
263 |
-
num_variations = st.selectbox(
|
264 |
-
"🔄 Email Variations",
|
265 |
-
[1, 2, 3],
|
266 |
-
index=1,
|
267 |
-
help="Number of email variations to generate per contact"
|
268 |
-
)
|
269 |
-
|
270 |
-
st.markdown("---")
|
271 |
-
|
272 |
-
# Model info
|
273 |
-
st.subheader("🤖 AI Model")
|
274 |
-
st.info("**Vicuna-7B GGUF**\n\nOptimized for quality cold email generation")
|
275 |
-
|
276 |
-
# File upload
|
277 |
-
st.subheader("📁 Upload Your Lead List")
|
278 |
-
uploaded_file = st.file_uploader(
|
279 |
-
"Choose a CSV file",
|
280 |
-
type=['csv'],
|
281 |
-
help="Upload a CSV file with columns: name, email, company, linkedin_url"
|
282 |
-
)
|
283 |
-
|
284 |
-
# Sample CSV download
|
285 |
-
col1, col2 = st.columns(2)
|
286 |
-
with col1:
|
287 |
-
if st.button("📥 Download Sample CSV"):
|
288 |
-
sample_data = {
|
289 |
-
'name': ['John Smith', 'Jane Doe', 'Mike Johnson'],
|
290 |
-
'email': ['[email protected]', '[email protected]', '[email protected]'],
|
291 |
-
'company': ['TechCorp Inc', 'StartupXYZ', 'Creative Agency'],
|
292 |
-
'linkedin_url': [
|
293 |
-
'https://linkedin.com/company/techcorp',
|
294 |
-
'https://linkedin.com/company/startupxyz',
|
295 |
-
'https://linkedin.com/company/creative-agency'
|
296 |
-
]
|
297 |
-
}
|
298 |
-
sample_df = pd.DataFrame(sample_data)
|
299 |
-
csv = sample_df.to_csv(index=False)
|
300 |
-
st.download_button(
|
301 |
-
"📄 sample_leads.csv",
|
302 |
-
csv,
|
303 |
-
"sample_leads.csv",
|
304 |
-
"text/csv"
|
305 |
-
)
|
306 |
-
|
307 |
-
# Initialize results variable
|
308 |
-
results = None
|
309 |
-
|
310 |
-
if uploaded_file is not None:
|
311 |
-
try:
|
312 |
-
# Load and validate CSV
|
313 |
-
df = pd.read_csv(uploaded_file)
|
314 |
-
|
315 |
-
st.success(f"✅ Loaded {len(df)} leads from CSV")
|
316 |
-
|
317 |
-
# Validate required columns
|
318 |
-
required_columns = ['name', 'email', 'company', 'linkedin_url']
|
319 |
-
missing_columns = [col for col in required_columns if col not in df.columns]
|
320 |
-
|
321 |
-
if missing_columns:
|
322 |
-
st.error(f"❌ Missing required columns: {missing_columns}")
|
323 |
-
st.info("Required columns: name, email, company, linkedin_url")
|
324 |
-
return
|
325 |
-
|
326 |
-
# Display preview
|
327 |
-
with st.expander("👀 Preview Data", expanded=True):
|
328 |
-
st.dataframe(df.head())
|
329 |
-
|
330 |
-
# Validate data
|
331 |
-
issues = []
|
332 |
-
if df['email'].isnull().any():
|
333 |
-
issues.append("Some emails are missing")
|
334 |
-
if df['linkedin_url'].isnull().any():
|
335 |
-
issues.append("Some LinkedIn URLs are missing")
|
336 |
-
|
337 |
-
if issues:
|
338 |
-
st.warning("⚠️ Data Issues Found:")
|
339 |
-
for issue in issues:
|
340 |
-
st.write(f"- {issue}")
|
341 |
-
|
342 |
-
if not st.checkbox("Continue anyway?"):
|
343 |
-
return
|
344 |
-
|
345 |
-
# Process button
|
346 |
-
if st.button("🚀 Generate Cold Emails", type="primary"):
|
347 |
-
if len(df) > 10:
|
348 |
-
st.warning("⚠️ Processing more than 10 leads may take several minutes. Consider processing in smaller batches.")
|
349 |
-
if not st.checkbox("I understand this may take a while"):
|
350 |
-
return
|
351 |
-
|
352 |
-
# Process the data
|
353 |
-
with st.spinner("🔄 Processing leads and generating emails..."):
|
354 |
-
results = process_csv_data(df, tone, creativity, num_variations)
|
355 |
-
|
356 |
-
if results:
|
357 |
-
st.success("✅ Email generation complete!")
|
358 |
-
|
359 |
-
# Display results immediately
|
360 |
-
st.markdown("---")
|
361 |
-
display_results(results, tone, creativity)
|
362 |
-
|
363 |
-
# Export functionality
|
364 |
-
st.markdown("---")
|
365 |
-
st.subheader("📤 Export Results")
|
366 |
-
|
367 |
-
export_df = export_to_csv(results)
|
368 |
-
if export_df is not None:
|
369 |
-
csv = export_df.to_csv(index=False)
|
370 |
-
st.download_button(
|
371 |
-
"📥 Download Results as CSV",
|
372 |
-
csv,
|
373 |
-
f"cold_emails_export_{datetime.now().strftime('%Y%m%d_%H%M%S')}.csv",
|
374 |
-
"text/csv",
|
375 |
-
help="Download all generated emails in CSV format"
|
376 |
-
)
|
377 |
-
|
378 |
-
st.info(f"📊 Ready to export {len(export_df)} successful email generations")
|
379 |
-
else:
|
380 |
-
st.error("❌ Failed to process leads. Please try again.")
|
381 |
-
|
382 |
-
except Exception as e:
|
383 |
-
st.error(f"❌ Error reading CSV file: {str(e)}")
|
384 |
-
st.info("Please ensure your CSV file has the correct format and encoding.")
|
385 |
-
|
386 |
-
# Footer
|
387 |
-
st.markdown("---")
|
388 |
-
st.markdown(
|
389 |
-
"""
|
390 |
-
<div style='text-align: center; color: #666;'>
|
391 |
-
<p>🚀 Cold Email Outreach Assistant | Built with Streamlit & Vicuna-7B</p>
|
392 |
-
<p>💡 Tip: Use specific, researched LinkedIn company URLs for best results</p>
|
393 |
-
</div>
|
394 |
-
""",
|
395 |
-
unsafe_allow_html=True
|
396 |
-
)
|
397 |
-
|
398 |
-
if __name__ == "__main__":
|
399 |
-
main()
|
|
|
1 |
import streamlit as st
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
|
|
|
3 |
st.set_page_config(
|
4 |
page_title="Cold Email Outreach Assistant",
|
5 |
page_icon="📧",
|
6 |
layout="wide"
|
7 |
)
|
8 |
|
9 |
+
st.title("📧 Cold Email Outreach Assistant")
|
10 |
+
st.write("🎉 Basic Streamlit app is working!")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
|
12 |
+
# Test file upload
|
13 |
+
uploaded_file = st.file_uploader("Test CSV Upload", type=['csv'])
|
14 |
+
if uploaded_file:
|
15 |
+
st.success("File upload works!")
|
|
|
|
|
|
|
|
|
|
|
16 |
|
17 |
+
st.write("If you can see this, the basic app is running correctly.")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|