AnishDev commited on
Commit
d3a8deb
Β·
1 Parent(s): d3bec7c

docker - setup

Browse files
Dockerfile CHANGED
@@ -22,7 +22,8 @@ RUN python -c "from sentence_transformers import SentenceTransformer; SentenceTr
22
  # Copy the application code into the container
23
  COPY app/ ./app/
24
 
25
- # HF Spaces requires port 7860
26
- EXPOSE 7860
 
27
 
28
- CMD ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "7860"]
 
22
  # Copy the application code into the container
23
  COPY app/ ./app/
24
 
25
+ # Default port 8000 for local/Docker. HF Spaces overrides with PORT=7860.
26
+ ENV PORT=8000
27
+ EXPOSE 8000
28
 
29
+ CMD ["sh", "-c", "uvicorn app.main:app --host 0.0.0.0 --port ${PORT}"]
sample_data/generate_sample_data.py ADDED
@@ -0,0 +1,551 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """
2
+ Generates sample_candidates.xlsx for testing the JD Candidate Matching System.
3
+
4
+ Run from the sample_data/ directory:
5
+ python generate_sample_data.py
6
+
7
+ Produces: sample_candidates.xlsx (15 candidates, mix of AI and Backend Engineers)
8
+
9
+ Column schema matches the expected XLSX format:
10
+ name, parsed_skills, parsed_summary, parsed_work_experience,
11
+ parsed_metadata_calculated_years_experience, engineer_type,
12
+ current_position, current_title, gen_ai_experience,
13
+ is_actively_or_passively_looking, parsed_metadata_most_recent_company,
14
+ notice_period
15
+ """
16
+ import openpyxl
17
+ from openpyxl.styles import Font, PatternFill, Alignment
18
+ from pathlib import Path
19
+
20
+
21
+ def we(company, role, duration, description):
22
+ """Format a single work experience entry as Python dict literal."""
23
+ desc = description.replace("'", "\\'")
24
+ return f"{{'company': '{company}', 'role': '{role}', 'duration': '{duration}', 'description': '{desc}'}}"
25
+
26
+
27
+ CANDIDATES = [
28
+ # ── AI Engineers ─────────────────────────────────────────────────────────
29
+ {
30
+ "id": "AI-001",
31
+ "name": "Priya Sharma",
32
+ "engineer_type": "AI Engineer",
33
+ "current_position": "Senior AI Engineer",
34
+ "current_title": "Senior AI Engineer",
35
+ "parsed_metadata_calculated_years_experience": 5.5,
36
+ "parsed_metadata_most_recent_company": "Google India",
37
+ "gen_ai_experience": "2+ years",
38
+ "is_actively_or_passively_looking": True,
39
+ "notice_period": "30 days",
40
+ "parsed_skills": (
41
+ "Python (Expert), LangChain (Advanced), RAG pipelines (Advanced), "
42
+ "PyTorch (Advanced), HuggingFace Transformers (Advanced), "
43
+ "FastAPI (Core), Vector Databases - Pinecone FAISS (Advanced), "
44
+ "LLM fine-tuning (Core), AWS SageMaker (Core), Docker (Core)"
45
+ ),
46
+ "parsed_summary": (
47
+ "AI Engineer with 5+ years building production ML systems. "
48
+ "Led RAG pipeline for Google Search quality team. "
49
+ "Published research on LLM evaluation metrics."
50
+ ),
51
+ "parsed_work_experience": ", ".join([
52
+ we("Google India", "Senior AI Engineer", "2.5 years",
53
+ "Led RAG pipeline design serving 10M queries/day. Fine-tuned LLaMA 3 on domain-specific corpora. "
54
+ "Built LangChain agents for automated QA. Used Pinecone for vector search. Deployed on GCP Vertex AI."),
55
+ we("Flipkart AI Labs", "ML Engineer", "2 years",
56
+ "Built recommendation models using PyTorch. Developed FastAPI inference service. "
57
+ "RAG-based product Q&A system using FAISS. LLM prompt engineering for catalog enrichment."),
58
+ we("Infosys AI Studio", "Junior ML Engineer", "1 year",
59
+ "Trained NLP classification models with HuggingFace. Deployed models on AWS SageMaker. "
60
+ "Experimented with GPT-3 embeddings for semantic search."),
61
+ ]),
62
+ },
63
+ {
64
+ "id": "AI-002",
65
+ "name": "Rahul Mehta",
66
+ "engineer_type": "AI Engineer",
67
+ "current_position": "AI Engineer",
68
+ "current_title": "AI Engineer",
69
+ "parsed_metadata_calculated_years_experience": 4.0,
70
+ "parsed_metadata_most_recent_company": "Swiggy",
71
+ "gen_ai_experience": "1.5 years - 2 years",
72
+ "is_actively_or_passively_looking": True,
73
+ "notice_period": "45 days",
74
+ "parsed_skills": (
75
+ "Python (Core), TensorFlow (Advanced), FastAPI (Core), "
76
+ "LangChain (Basic), OpenAI API (Core), "
77
+ "PostgreSQL (Core), Redis (Basic), Docker (Basic), AWS (Basic)"
78
+ ),
79
+ "parsed_summary": (
80
+ "AI Engineer at Swiggy building demand forecasting and LLM-powered "
81
+ "chatbot systems. Experience with LLM APIs and RAG architectures."
82
+ ),
83
+ "parsed_work_experience": ", ".join([
84
+ we("Swiggy", "AI Engineer", "2 years",
85
+ "Built LLM-powered customer support chatbot using OpenAI GPT-4 and LangChain. "
86
+ "Integrated RAG with Redis vector store. Developed FastAPI endpoints for real-time inference. "
87
+ "Demand forecasting using TensorFlow time-series models."),
88
+ we("Ola Electric", "Data Scientist", "2 years",
89
+ "Built anomaly detection models for battery management. "
90
+ "Developed Python ETL pipelines and SQL dashboards. "
91
+ "First LLM experiment: GPT-3 for log summarisation."),
92
+ ]),
93
+ },
94
+ {
95
+ "id": "AI-003",
96
+ "name": "Ananya Patel",
97
+ "engineer_type": "AI Engineer",
98
+ "current_position": "Principal AI Engineer",
99
+ "current_title": "Principal AI Engineer",
100
+ "parsed_metadata_calculated_years_experience": 7.0,
101
+ "parsed_metadata_most_recent_company": "Microsoft India",
102
+ "gen_ai_experience": "2+ years",
103
+ "is_actively_or_passively_looking": False,
104
+ "notice_period": "60 days",
105
+ "parsed_skills": (
106
+ "Python (Expert), LangChain (Expert), LangGraph (Advanced), "
107
+ "RAG (Expert), Vector Databases (Expert), LLM fine-tuning (Advanced), "
108
+ "Azure OpenAI (Expert), HuggingFace (Advanced), FastAPI (Advanced), "
109
+ "Kubernetes (Core), MLflow (Core)"
110
+ ),
111
+ "parsed_summary": (
112
+ "Principal AI Engineer at Microsoft with 7 years in ML/AI. "
113
+ "Architect of Azure OpenAI enterprise integrations. "
114
+ "Expert in multi-agent LangGraph systems and production RAG."
115
+ ),
116
+ "parsed_work_experience": ", ".join([
117
+ we("Microsoft India", "Principal AI Engineer", "3 years",
118
+ "Architected enterprise RAG systems using Azure OpenAI and Cognitive Search. "
119
+ "Built multi-agent LangGraph workflows for document processing. "
120
+ "Led team of 8 engineers. LLM fine-tuning on proprietary datasets. "
121
+ "Pinecone and Weaviate for vector search at scale."),
122
+ we("Samsung Research", "Senior ML Engineer", "2.5 years",
123
+ "Built on-device NLP models using HuggingFace and PyTorch. "
124
+ "Generative AI for Bixby voice assistant. "
125
+ "Fine-tuned LLaMA models for multilingual tasks."),
126
+ we("Mu Sigma", "ML Engineer", "1.5 years",
127
+ "Statistical ML models for retail analytics. Python, scikit-learn, FastAPI."),
128
+ ]),
129
+ },
130
+ {
131
+ "id": "AI-004",
132
+ "name": "Sneha Iyer",
133
+ "engineer_type": "AI Engineer",
134
+ "current_position": "ML Engineer",
135
+ "current_title": "ML Engineer",
136
+ "parsed_metadata_calculated_years_experience": 3.5,
137
+ "parsed_metadata_most_recent_company": "Freshworks",
138
+ "gen_ai_experience": "6 months - 1 year",
139
+ "is_actively_or_passively_looking": True,
140
+ "notice_period": "30 days",
141
+ "parsed_skills": (
142
+ "Python (Core), scikit-learn (Advanced), FastAPI (Basic), "
143
+ "LangChain (Basic), OpenAI API (Basic), "
144
+ "SQL (Core), pandas (Core), Docker (Basic)"
145
+ ),
146
+ "parsed_summary": (
147
+ "ML Engineer at Freshworks. Recently started exploring LLMs and GenAI. "
148
+ "Strong classical ML background, transitioning to GenAI workloads."
149
+ ),
150
+ "parsed_work_experience": ", ".join([
151
+ we("Freshworks", "ML Engineer", "2 years",
152
+ "Built NLP models for customer sentiment analysis using scikit-learn. "
153
+ "Developed FastAPI endpoints for ML inference. "
154
+ "Started exploring LangChain for RAG-based knowledge base. "
155
+ "Used OpenAI API for text classification experiments."),
156
+ we("Tata Consultancy Services", "Software Engineer", "1.5 years",
157
+ "Developed Python data pipelines. SQL reporting. Some ML model training."),
158
+ ]),
159
+ },
160
+ {
161
+ "id": "AI-005",
162
+ "name": "Vikram Nair",
163
+ "engineer_type": "AI Engineer",
164
+ "current_position": "Software Engineer",
165
+ "current_title": "Software Engineer",
166
+ "parsed_metadata_calculated_years_experience": 2.5,
167
+ "parsed_metadata_most_recent_company": "Wipro",
168
+ "gen_ai_experience": "only internship/project experience",
169
+ "is_actively_or_passively_looking": True,
170
+ "notice_period": "15 days",
171
+ "parsed_skills": (
172
+ "Python (Basic), scikit-learn (Basic), TensorFlow (Basic), "
173
+ "pandas (Core), SQL (Core), REST APIs (Basic)"
174
+ ),
175
+ "parsed_summary": (
176
+ "Software Engineer pivoting to AI/ML. "
177
+ "Completed LLM project during internship. "
178
+ "Eager to grow in GenAI space."
179
+ ),
180
+ "parsed_work_experience": ", ".join([
181
+ we("Wipro", "Software Engineer", "2.5 years",
182
+ "Backend Java development. REST APIs. Some Python scripting for data tasks. "
183
+ "Built a small text classifier using scikit-learn as a side project."),
184
+ ]),
185
+ },
186
+ {
187
+ "id": "AI-006",
188
+ "name": "Divya Krishnan",
189
+ "engineer_type": "AI Engineer",
190
+ "current_position": "Research Engineer",
191
+ "current_title": "AI Research Engineer",
192
+ "parsed_metadata_calculated_years_experience": 4.5,
193
+ "parsed_metadata_most_recent_company": "IIT Bangalore AI Lab",
194
+ "gen_ai_experience": "2+ years",
195
+ "is_actively_or_passively_looking": True,
196
+ "notice_period": "30 days",
197
+ "parsed_skills": (
198
+ "Python (Expert), PyTorch (Expert), HuggingFace Transformers (Expert), "
199
+ "LLM fine-tuning (Advanced), RLHF (Core), "
200
+ "FAISS (Advanced), LangChain (Core), FastAPI (Core), "
201
+ "CUDA (Basic), MLflow (Core)"
202
+ ),
203
+ "parsed_summary": (
204
+ "Research Engineer specializing in LLM fine-tuning and RLHF. "
205
+ "Published 2 papers on instruction-tuned models. "
206
+ "Strong PyTorch and HuggingFace expertise."
207
+ ),
208
+ "parsed_work_experience": ", ".join([
209
+ we("IIT Bangalore AI Lab", "Research Engineer", "2 years",
210
+ "Fine-tuned LLaMA 2 and Mistral models using RLHF and LoRA. "
211
+ "Built RAG evaluation benchmarks using FAISS and LangChain. "
212
+ "Published research on hallucination mitigation in LLMs. "
213
+ "Generative AI for code generation tasks."),
214
+ we("Anthropic (Intern)", "AI Safety Research Intern", "6 months",
215
+ "Red-teaming Claude models. Prompt injection research. Constitutional AI experiments."),
216
+ we("Samsung Research", "ML Engineer", "2 years",
217
+ "On-device transformer models with HuggingFace. MLflow for experiment tracking. "
218
+ "FastAPI for model serving. RAG prototype using FAISS."),
219
+ ]),
220
+ },
221
+ {
222
+ "id": "AI-007",
223
+ "name": "Arjun Bose",
224
+ "engineer_type": "AI Engineer",
225
+ "current_position": "Data Scientist",
226
+ "current_title": "Senior Data Scientist",
227
+ "parsed_metadata_calculated_years_experience": 6.0,
228
+ "parsed_metadata_most_recent_company": "Paytm",
229
+ "gen_ai_experience": "1 to 1.5 years",
230
+ "is_actively_or_passively_looking": False,
231
+ "notice_period": "90 days",
232
+ "parsed_skills": (
233
+ "Python (Expert), XGBoost (Expert), TensorFlow (Core), "
234
+ "SQL (Expert), Spark (Core), AWS (Core), "
235
+ "OpenAI API (Basic), LangChain (Basic)"
236
+ ),
237
+ "parsed_summary": (
238
+ "Senior Data Scientist at Paytm with strong classical ML background. "
239
+ "Recently began integrating LLMs into fraud detection pipeline. "
240
+ "More DS than AI Engineer profile."
241
+ ),
242
+ "parsed_work_experience": ", ".join([
243
+ we("Paytm", "Senior Data Scientist", "3 years",
244
+ "Built fraud detection models with XGBoost. A/B testing frameworks. "
245
+ "Recently added LLM-based transaction explanation using OpenAI API. "
246
+ "LangChain for internal audit Q&A chatbot."),
247
+ we("HDFC Bank", "Data Analyst", "3 years",
248
+ "SQL reporting and Python data pipelines. Credit risk models."),
249
+ ]),
250
+ },
251
+ # ── Backend Engineers ─────────────────────────────────────────────────────
252
+ {
253
+ "id": "BE-001",
254
+ "name": "Kiran Kumar",
255
+ "engineer_type": "Backend Engineer",
256
+ "current_position": "Senior Backend Engineer",
257
+ "current_title": "Senior Backend Engineer",
258
+ "parsed_metadata_calculated_years_experience": 6.0,
259
+ "parsed_metadata_most_recent_company": "Razorpay",
260
+ "gen_ai_experience": "i do not have gen ai experience",
261
+ "is_actively_or_passively_looking": True,
262
+ "notice_period": "30 days",
263
+ "parsed_skills": (
264
+ "Node.js (Expert), Express.js (Expert), AWS (Advanced), "
265
+ "PostgreSQL (Advanced), Redis (Advanced), "
266
+ "Microservices (Advanced), Docker (Advanced), "
267
+ "Kubernetes (Core), REST APIs (Expert), GraphQL (Core)"
268
+ ),
269
+ "parsed_summary": (
270
+ "Senior Backend Engineer at Razorpay. 6 years building high-throughput "
271
+ "payment APIs. Expert in Node.js microservices on AWS."
272
+ ),
273
+ "parsed_work_experience": ", ".join([
274
+ we("Razorpay", "Senior Backend Engineer", "3 years",
275
+ "Led backend for payment gateway handling 50K TPS. Node.js microservices on AWS ECS. "
276
+ "PostgreSQL + Redis for transaction state. Built GraphQL API layer. "
277
+ "Kafka for event streaming between services."),
278
+ we("PhonePe", "Backend Engineer", "2 years",
279
+ "Node.js Express APIs for UPI payment flows. "
280
+ "Redis caching for session management. PostgreSQL sharding. Docker deployment."),
281
+ we("Cognizant", "Software Engineer", "1 year",
282
+ "Java Spring Boot backend. REST API development. MySQL. Agile team."),
283
+ ]),
284
+ },
285
+ {
286
+ "id": "BE-002",
287
+ "name": "Meera Joshi",
288
+ "engineer_type": "Backend Engineer",
289
+ "current_position": "Backend Engineer",
290
+ "current_title": "Backend Engineer",
291
+ "parsed_metadata_calculated_years_experience": 4.0,
292
+ "parsed_metadata_most_recent_company": "Zomato",
293
+ "gen_ai_experience": None,
294
+ "is_actively_or_passively_looking": True,
295
+ "notice_period": "45 days",
296
+ "parsed_skills": (
297
+ "Node.js (Advanced), Express.js (Advanced), MongoDB (Advanced), "
298
+ "PostgreSQL (Core), Redis (Core), AWS (Core), "
299
+ "Docker (Core), REST APIs (Advanced), Microservices (Core)"
300
+ ),
301
+ "parsed_summary": (
302
+ "Backend Engineer at Zomato. Strong Node.js and Express expertise. "
303
+ "Built food delivery order management APIs serving millions of users."
304
+ ),
305
+ "parsed_work_experience": ", ".join([
306
+ we("Zomato", "Backend Engineer", "2.5 years",
307
+ "Built order management microservices in Node.js/Express. "
308
+ "MongoDB for order storage. Redis for caching menu data. "
309
+ "AWS Lambda for async order events. REST APIs consumed by mobile apps."),
310
+ we("MakeMyTrip", "Junior Backend Engineer", "1.5 years",
311
+ "Node.js APIs for hotel booking. PostgreSQL. Docker containers."),
312
+ ]),
313
+ },
314
+ {
315
+ "id": "BE-003",
316
+ "name": "Arjun Singh",
317
+ "engineer_type": "Backend Engineer",
318
+ "current_position": "Tech Lead",
319
+ "current_title": "Backend Tech Lead",
320
+ "parsed_metadata_calculated_years_experience": 8.0,
321
+ "parsed_metadata_most_recent_company": "Zerodha",
322
+ "gen_ai_experience": "i do not have gen ai experience",
323
+ "is_actively_or_passively_looking": False,
324
+ "notice_period": "90 days",
325
+ "parsed_skills": (
326
+ "Node.js (Expert), Express.js (Expert), AWS (Expert), "
327
+ "PostgreSQL (Expert), DynamoDB (Advanced), Redis (Advanced), "
328
+ "Microservices (Expert), Docker (Expert), Kubernetes (Advanced), "
329
+ "REST APIs (Expert), System Design (Expert)"
330
+ ),
331
+ "parsed_summary": (
332
+ "Tech Lead at Zerodha (India's largest stock broker). "
333
+ "8 years of backend experience. Led architecture of trading platform "
334
+ "backend serving 10M users."
335
+ ),
336
+ "parsed_work_experience": ", ".join([
337
+ we("Zerodha", "Backend Tech Lead", "4 years",
338
+ "Architected Node.js microservices for stock trading platform. "
339
+ "AWS EKS for container orchestration. DynamoDB for order book. "
340
+ "PostgreSQL for account management. Redis Cluster for real-time positions. "
341
+ "Led team of 12 backend engineers."),
342
+ we("CRED", "Senior Backend Engineer", "2.5 years",
343
+ "High-performance REST APIs in Node.js/Express. AWS + Kubernetes deployment. "
344
+ "PostgreSQL with read replicas. Redis caching."),
345
+ we("Infosys", "Software Engineer", "1.5 years",
346
+ "Java Spring Boot. REST APIs. MySQL. Agile development."),
347
+ ]),
348
+ },
349
+ {
350
+ "id": "BE-004",
351
+ "name": "Kavya Reddy",
352
+ "engineer_type": "Backend Engineer",
353
+ "current_position": "Backend Engineer",
354
+ "current_title": "Software Engineer - Backend",
355
+ "parsed_metadata_calculated_years_experience": 3.0,
356
+ "parsed_metadata_most_recent_company": "OYO Rooms",
357
+ "gen_ai_experience": None,
358
+ "is_actively_or_passively_looking": True,
359
+ "notice_period": "30 days",
360
+ "parsed_skills": (
361
+ "Node.js (Core), Express.js (Advanced), MongoDB (Advanced), "
362
+ "AWS (Basic), Docker (Basic), REST APIs (Advanced)"
363
+ ),
364
+ "parsed_summary": (
365
+ "Backend Engineer at OYO with 3 years Node.js experience. "
366
+ "Solid Express and MongoDB skills. Limited cloud exposure."
367
+ ),
368
+ "parsed_work_experience": ", ".join([
369
+ we("OYO Rooms", "Backend Engineer", "2 years",
370
+ "Node.js/Express APIs for hotel inventory management. "
371
+ "MongoDB for property data. REST APIs for mobile apps. "
372
+ "Basic AWS EC2 deployment."),
373
+ we("Wipro", "Junior Developer", "1 year",
374
+ "Python and Node.js scripting. REST API testing. QA tasks."),
375
+ ]),
376
+ },
377
+ {
378
+ "id": "BE-005",
379
+ "name": "Dev Agarwal",
380
+ "engineer_type": "Backend Engineer",
381
+ "current_position": "Junior Backend Developer",
382
+ "current_title": "Backend Developer",
383
+ "parsed_metadata_calculated_years_experience": 1.5,
384
+ "parsed_metadata_most_recent_company": "Startup XYZ",
385
+ "gen_ai_experience": None,
386
+ "is_actively_or_passively_looking": True,
387
+ "notice_period": "15 days",
388
+ "parsed_skills": (
389
+ "Node.js (Basic), Express.js (Basic), MongoDB (Basic), "
390
+ "REST APIs (Basic), JavaScript (Core)"
391
+ ),
392
+ "parsed_summary": (
393
+ "Junior backend developer with 1.5 years experience. "
394
+ "Learning Node.js and Express. Early in career."
395
+ ),
396
+ "parsed_work_experience": ", ".join([
397
+ we("Startup XYZ", "Junior Backend Developer", "1.5 years",
398
+ "Node.js Express REST APIs for e-commerce app. MongoDB CRUD operations. "
399
+ "Basic deployment on Heroku."),
400
+ ]),
401
+ },
402
+ # ── Edge Cases ────────────────────────────────────────────────────────────
403
+ {
404
+ "id": "EDGE-001",
405
+ "name": "Rohan Verma",
406
+ "engineer_type": "AI Engineer",
407
+ "current_position": "Fresh Graduate",
408
+ "current_title": None,
409
+ "parsed_metadata_calculated_years_experience": None, # years_unknown
410
+ "parsed_metadata_most_recent_company": None,
411
+ "gen_ai_experience": "only internship/project experience",
412
+ "is_actively_or_passively_looking": True,
413
+ "notice_period": "Immediate",
414
+ "parsed_skills": (
415
+ "Python (Basic), TensorFlow (Basic), scikit-learn (Basic), "
416
+ "LangChain (Basic), SQL (Basic)"
417
+ ),
418
+ "parsed_summary": (
419
+ "Recent CS graduate. Built a RAG chatbot as final year project using LangChain. "
420
+ "Eager to start career in AI Engineering."
421
+ ),
422
+ "parsed_work_experience": ", ".join([
423
+ we("University Project", "AI Intern", "6 months",
424
+ "Built RAG-based document Q&A using LangChain and FAISS. "
425
+ "Experimented with LLM prompt engineering. Python, FastAPI for serving."),
426
+ ]),
427
+ },
428
+ {
429
+ "id": "EDGE-002",
430
+ "name": "Pooja Gupta",
431
+ "engineer_type": "Frontend Engineer", # wrong type for AI JD
432
+ "current_position": "Frontend Developer",
433
+ "current_title": "React Developer",
434
+ "parsed_metadata_calculated_years_experience": 4.0,
435
+ "parsed_metadata_most_recent_company": "Accenture",
436
+ "gen_ai_experience": "i do not have gen ai experience",
437
+ "is_actively_or_passively_looking": False,
438
+ "notice_period": "30 days",
439
+ "parsed_skills": (
440
+ "React (Expert), JavaScript (Expert), TypeScript (Advanced), "
441
+ "Next.js (Advanced), CSS (Expert), REST APIs (Core)"
442
+ ),
443
+ "parsed_summary": (
444
+ "Senior Frontend Developer specializing in React and Next.js. "
445
+ "No backend or AI experience."
446
+ ),
447
+ "parsed_work_experience": ", ".join([
448
+ we("Accenture", "Frontend Developer", "4 years",
449
+ "React and Next.js UI development. TypeScript. REST API integration. CSS animations."),
450
+ ]),
451
+ },
452
+ {
453
+ "id": "EDGE-003",
454
+ "name": "Siddharth Rao",
455
+ "engineer_type": "AI Engineer",
456
+ "current_position": "AI Engineer",
457
+ "current_title": "AI Engineer",
458
+ "parsed_metadata_calculated_years_experience": 3.0,
459
+ "parsed_metadata_most_recent_company": "TCS",
460
+ "gen_ai_experience": "2+ years", # claims 2+ years but work history won't support it
461
+ "is_actively_or_passively_looking": True,
462
+ "notice_period": "60 days",
463
+ "parsed_skills": (
464
+ "Python (Core), SQL (Core), pandas (Core), "
465
+ "Excel (Advanced), PowerPoint (Advanced)"
466
+ ),
467
+ "parsed_summary": (
468
+ "AI Engineer at TCS with self-reported 2+ years GenAI experience. "
469
+ "Work history reveals primarily data analyst tasks, not AI engineering."
470
+ ),
471
+ "parsed_work_experience": ", ".join([
472
+ we("TCS", "Business Analyst", "3 years",
473
+ "Data analysis using SQL and Excel. PowerPoint presentations for clients. "
474
+ "Python scripting for data cleanup. No LLM or ML model work."),
475
+ ]),
476
+ },
477
+ ]
478
+
479
+
480
+ def build_workbook():
481
+ wb = openpyxl.Workbook()
482
+ ws = wb.active
483
+ ws.title = "Candidates"
484
+
485
+ headers = [
486
+ "id", "name", "engineer_type", "current_position", "current_title",
487
+ "parsed_metadata_calculated_years_experience",
488
+ "parsed_metadata_most_recent_company",
489
+ "gen_ai_experience", "is_actively_or_passively_looking", "notice_period",
490
+ "parsed_skills", "parsed_summary", "parsed_work_experience",
491
+ ]
492
+
493
+ # Header row styling
494
+ header_fill = PatternFill("solid", fgColor="1F4E79")
495
+ header_font = Font(bold=True, color="FFFFFF", size=11)
496
+ for col_idx, header in enumerate(headers, start=1):
497
+ cell = ws.cell(row=1, column=col_idx, value=header)
498
+ cell.fill = header_fill
499
+ cell.font = header_font
500
+ cell.alignment = Alignment(horizontal="center", vertical="center")
501
+
502
+ # Data rows
503
+ row_fill_a = PatternFill("solid", fgColor="EBF3FB")
504
+ row_fill_b = PatternFill("solid", fgColor="FFFFFF")
505
+ wrap = Alignment(wrap_text=True, vertical="top")
506
+
507
+ for row_idx, cand in enumerate(CANDIDATES, start=2):
508
+ fill = row_fill_a if row_idx % 2 == 0 else row_fill_b
509
+ values = [
510
+ cand.get("id"),
511
+ cand.get("name"),
512
+ cand.get("engineer_type"),
513
+ cand.get("current_position"),
514
+ cand.get("current_title"),
515
+ cand.get("parsed_metadata_calculated_years_experience"),
516
+ cand.get("parsed_metadata_most_recent_company"),
517
+ cand.get("gen_ai_experience"),
518
+ cand.get("is_actively_or_passively_looking"),
519
+ cand.get("notice_period"),
520
+ cand.get("parsed_skills"),
521
+ cand.get("parsed_summary"),
522
+ cand.get("parsed_work_experience"),
523
+ ]
524
+ for col_idx, value in enumerate(values, start=1):
525
+ cell = ws.cell(row=row_idx, column=col_idx, value=value)
526
+ cell.fill = fill
527
+ cell.alignment = wrap
528
+
529
+ # Column widths
530
+ widths = [12, 20, 18, 22, 22, 10, 22, 20, 10, 12, 60, 50, 120]
531
+ for col_idx, width in enumerate(widths, start=1):
532
+ ws.column_dimensions[ws.cell(row=1, column=col_idx).column_letter].width = width
533
+
534
+ ws.row_dimensions[1].height = 20
535
+ ws.freeze_panes = "A2"
536
+
537
+ return wb
538
+
539
+
540
+ if __name__ == "__main__":
541
+ out_path = Path(__file__).parent / "sample_candidates.xlsx"
542
+ wb = build_workbook()
543
+ wb.save(out_path)
544
+ print(f"Generated: {out_path}")
545
+ print(f"Total candidates: {len(CANDIDATES)}")
546
+ types = {}
547
+ for c in CANDIDATES:
548
+ t = c["engineer_type"]
549
+ types[t] = types.get(t, 0) + 1
550
+ for t, n in types.items():
551
+ print(f" {t}: {n}")
sample_data/sample_candidates.xlsx ADDED
Binary file (10.2 kB). View file
 
sample_data/sample_job_description.md ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Sample Job Description β€” AI Engineer (Used for Testing)
2
+
3
+ This file represents the kind of JD the system is designed to parse.
4
+ To test the system, convert this to a PDF and upload it via the UI.
5
+
6
+ ---
7
+
8
+ ## Job Title: AI Engineer
9
+
10
+ **Company:** Wissen Technology
11
+ **Location:** Bangalore, India (Hybrid)
12
+ **Experience Required:** 3+ years
13
+
14
+ ---
15
+
16
+ ## About the Role
17
+
18
+ We are looking for a skilled AI Engineer to join our growing AI/ML team.
19
+ You will design, build, and deploy production-grade AI systems β€” including
20
+ LLM pipelines, RAG architectures, and real-time inference services.
21
+
22
+ ---
23
+
24
+ ## Responsibilities
25
+
26
+ - Design and implement RAG (Retrieval-Augmented Generation) pipelines for enterprise clients
27
+ - Fine-tune and evaluate large language models (GPT-4, Claude, LLaMA)
28
+ - Build vector search systems using FAISS, Pinecone, or Weaviate
29
+ - Develop and maintain ML model serving infrastructure (FastAPI, Triton)
30
+ - Collaborate with backend teams to integrate AI APIs into production systems
31
+ - Write clean, tested, production-ready Python code
32
+ - Mentor junior team members and conduct code reviews
33
+
34
+ ---
35
+
36
+ ## Required Skills
37
+
38
+ - Python (Core) β€” 3+ years, production experience
39
+ - LLM / Large Language Models β€” experience with GPT, Claude, or Gemini APIs
40
+ - RAG (Retrieval-Augmented Generation) β€” pipeline design and evaluation
41
+ - LangChain or LangGraph β€” agent and chain orchestration
42
+ - Vector Databases β€” Pinecone, Weaviate, FAISS, or Chroma
43
+ - FastAPI β€” building and deploying inference APIs
44
+ - Machine Learning β€” model training, evaluation, deployment
45
+
46
+ ## Preferred Skills
47
+
48
+ - PyTorch β€” deep learning model training
49
+ - HuggingFace Transformers β€” model loading, fine-tuning
50
+ - Docker β€” containerized deployment
51
+ - AWS or GCP β€” cloud deployment, SageMaker, Vertex AI
52
+ - MLflow or W&B β€” experiment tracking
53
+
54
+ ---
55
+
56
+ ## GenAI Experience
57
+
58
+ Minimum 1 year of hands-on GenAI experience required.
59
+ Candidates with "2+ years" of production GenAI work preferred.
60
+
61
+ ---
62
+
63
+ ## What We Offer
64
+
65
+ - Competitive salary + ESOP
66
+ - Work on cutting-edge AI projects
67
+ - Hybrid work model
68
+ - Learning budget for certifications and conferences