Types of companies hiring DL talent in Gurugram:
- Global capability centres (GCCs)
- Product companies with AI teams
- AI-first startups (GenAI, vision, speech)
- Consulting firms building AI practices
- Fintech, healthtech, and adtech companies
- Edtech and e-commerce platforms

Most roles are 10-30 min from
Gurugram Metro and Rapid Metro stations.

Common entry-level DL roles in Gurugram:
- Junior ML Engineer
- AI Engineer (LLM / GenAI)
- Data Scientist (DL-focused)
- Computer Vision Engineer (junior)
- MLOps Associate
- AI Research Assistant

Most roles ask for:
- Python
- PyTorch or TensorFlow
- One strong DL project on GitHub
- SQL and cloud basics

# PyTorch — a tiny image classifier
import torch
import torch.nn as nn
import torch.optim as optim

class TinyNet(nn.Module):
    def __init__(self):
        super().__init__()
        self.net = nn.Sequential(
            nn.Flatten(),
            nn.Linear(28*28, 128),
            nn.ReLU(),
            nn.Linear(128, 10),
        )
    def forward(self, x):
        return self.net(x)

model = TinyNet()
loss_fn = nn.CrossEntropyLoss()
opt = optim.Adam(model.parameters(), lr=1e-3)

# Training loop
for epoch in range(5):
    for xb, yb in train_loader:
        logits = model(xb)
        loss = loss_fn(logits, yb)

        opt.zero_grad()
        loss.backward()
        opt.step()

    print(f"Epoch {epoch+1} loss: {loss.item():.4f}")

# Hugging Face — run an LLM locally
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

model_id = "gpt2"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

prompt = "Explain deep learning in one sentence:"
inputs = tokenizer(prompt, return_tensors="pt")

with torch.no_grad():
    output = model.generate(
        **inputs,
        max_new_tokens=80,
        do_sample=True,
        temperature=0.7,
    )

text = tokenizer.decode(output[0], skip_special_tokens=True)
print(text)

Project 1 — Sentiment classifier with Hugging Face
1. Load a public reviews dataset
2. Fine-tune DistilBERT for sentiment
3. Evaluate on a held-out split
4. Serve it via FastAPI
5. Dockerise and deploy to a free host
6. README with metrics and demo link

Skills shown: PyTorch, transformers,
              evaluation, deployment,
              Docker, FastAPI.

Project 2 — RAG chatbot on your notes
1. Collect PDFs (your own notes/docs)
2. Chunk and embed with an embedding model
3. Store in a vector DB (Chroma or FAISS)
4. Retrieve top-k chunks for a query
5. Generate an answer using an LLM
6. Add a simple Flask or Streamlit UI

Skills shown: embeddings, RAG,
              LLM prompting, vector DBs,
              API design, UI basics.

Weekly study plan for students (12-15 hrs):
Mon 1.5h — Theory + notes
Tue 1.5h — Coding practice
Wed 1.5h — Theory + notes
Thu 1.5h — Coding practice
Fri 1.5h — Weekly review
Sat 4h   — Project work
Sun 2h   — Reading + rest

Consistency beats intensity.
Join a Gurugram AI meetup once a month.

Portfolio checklist for Gurugram hiring:
- GitHub profile with clean READMEs
- 2 DL/GenAI projects (one classic, one LLM)
- Each project: problem, approach, results
- One deployed project with a demo link
- LinkedIn with "Open to Work" + keywords
- A one-page resume tailored to DL roles

Show the projects in interviews —
talk through decisions, not just outputs.