Skills deep-dive

Data Science & ML

Section 1 / The Floor

The minimum skill expectations for the role

7MUST-HAVE SKILLS
Skills · ranked by prevalence
01Python77.7%the whole stack speaks it
02PyTorch43.9%the default deep learning entry
03Model Deployment35.1%shipping beats training
04Scikit-learn27.0%classical models still do the work
05MLOps21.6%models must survive production
06Pandas17.6%where every dataset gets handled
07NumPy16.9%the numeric floor under everything
Floor skills derived from validated fresher-accessible postings across Indian job boards for the data science & ml profile.
Section 2 / Pick Your Track

Where to specialize once you're in

01

CLASSICAL ML

← THE STATS CORE
FA share33.3%of fresher postings
Entry pay4to ₹11LPA · min to typical · profile-wide
On top of the floor
TensorFlowStatisticsFeature EngineeringModel Evaluation

MNC and GCC led, with services firms, product companies and a smaller-employer tail.

02

GENAI / LLM

← THE EMERGING TRACK
FA share22.7%of fresher postings
Entry pay4to ₹11LPA · min to typical · profile-wide
On top of the floor
AWSLangChainAzure

MNC-led, with established mid-size firms, services shops and a little elite tech.

Thin, fast-moving sample, so the GenAI picture will keep shifting.
03

COMPUTER VISION

← THE VISION NICHE
FA share22%of fresher postings
Entry pay4to ₹11LPA · min to typical · profile-wide
On top of the floor
Computer Vision AlgorithmsImage ProcessingOpenCVTensorFlowVideo Analytics

Broadly spread across MNCs, startups, mid-size firms and services, with no clear leader.

Small sample, so the vision niche's numbers are indicative only.
04

NLP

← THE PREMIUM OUTLIER
FA share14.7%of fresher postings
Entry pay4to ₹11LPA · min to typical · profile-wide
On top of the floor
Distributed SystemsSystem DesignInformation Retrieval

The most MAANG-weighted DS track, though on very few postings, with domestic and MNC roles behind.

Very thin sample, so the high pay rests on few postings.
05

SPEECH / AUDIO

← THE AUDIO NICHE
FA share4%of fresher postings
Entry pay4to ₹11LPA · min to typical · profile-wide
On top of the floor
Speech ProcessingTransformersModel EvaluationC/C++Reinforcement Learning

MAANG-weighted with MNC support, on a handful of specialist postings.

A handful of postings, so every figure is illustrative only.
Track shares and entry-level salary ranges (minimum to mean) derived from validated fresher-accessible postings across Indian job boards for the data science & ml profile.
Section 3 / The Skill Arc

The skills that grow with your seniority

A

Senior skew - what skills grow more important with seniority.

Rising · FA → SR
01MLOps
FA21.6%
SR39.2%
+17.6 pp
02Azure
FA20.9%
SR34.6%
+13.7 pp
03GCP
FA18.9%
SR32.3%
+13.4 pp
04Spark
FA4.7%
SR16.5%
+11.8 pp
05SQL
FA17.6%
SR28.1%
+10.5 pp
06Kubernetes
FA10.8%
SR20.4%
+9.6 pp
07TensorFlow
FA37.2%
SR46.2%
+9.0 pp
08Data Pipelines
FA12.8%
SR20.0%
+7.2 pp
09AWS
FA29.1%
SR36.2%
+7.1 pp
Fading · FA → SR
01Algorithms
FA35.8%
SR16.2%
-19.7 pp
02Computer Vision Algorithms
FA27.7%
SR13.5%
-14.2 pp
03Data Structures
FA23.6%
SR10.4%
-13.3 pp
04Image Processing
FA12.8%
SR3.1%
-9.8 pp
05Git
FA24.3%
SR14.6%
-9.7 pp
B

Portability - what skills remain valued in other profiles.

Portable · cross-profile spread (out of 15)
01SQL
12/15
Data Analytics & BI64%Data Engineering58%Fullstack Development42%
02Java
12/15
Backend Development59%QA & Testing42%Fullstack Development39%
03Python
11/15
AI & LLM Applications76%Data Engineering73%DevOps & Platform Engineering54%
04AWS
10/15
DevOps & Platform Engineering44%Data Engineering44%Fullstack Development41%
05Docker
8/15
Backend Development28%DevOps & Platform Engineering25%AI & LLM Applications24%
06Kubernetes
8/15
DevOps & Platform Engineering35%Security Engineering27%Backend Development22%
07Azure
7/15
Data Engineering30%AI & LLM Applications28%DevOps & Platform Engineering28%
08GCP
7/15
AI & LLM Applications26%Security Engineering25%DevOps & Platform Engineering22%
09Distributed Systems
7/15
Backend Development20%DevOps & Platform Engineering16%AI & LLM Applications14%
Locked · stays inside the profile
01Model Deployment1/15
02Computer Vision Algorithms1/15
03Scikit-learn1/15
04Statistics1/15
05Pandas1/15
06PyTorch2/15
07TensorFlow2/15
08Model Evaluation2/15
09MLOps2/15
10Model Monitoring2/15
Skill arc derived from validated postings across Indian job boards for the data science & ml profile, comparing fresher-accessible and senior prevalence. Cross-profile spread computed at a one-in-ten prevalence threshold across all 15 profiles.
Section 4 / Targeting your Company Pool

Which skills each employer tier wants

MAANG and Tier-1 Global Tech

n = 113usable
CompaniesGoogleMetaAmazonMicrosoft ResearchAnthropic
Skill skew · MAANG and Tier-1 Global Tech vs IT Services & BPO
% in MAANGSkill% in IT ServicesSkew
56.6%
Distributed Systems
8.0%
+48.6 pp
38.1%
Algorithms
2.0%
+36.1 pp
35.4%
Data Structures
0.0%
+35.4 pp
31.0%
C/C++
0.0%
+31.0 pp
28.3%
Information Retrieval
0.0%
+28.3 pp
30.1%
UI Design
2.0%
+28.1 pp
42.5%
Code Reviews
18.0%
+24.5 pp
25.7%
Java
2.0%
+23.7 pp
27.4%
Reinforcement Learning
4.0%
+23.4 pp

IT Services & BPO

n = 50borderline
CompaniesTCSInfosysWiproHCLCognizant
Skill skew · MAANG and Tier-1 Global Tech vs IT Services & BPO
% in MAANGSkill% in IT ServicesSkew
8.8%
PyTorch
52.0%
-43.2 pp
13.3%
TensorFlow
56.0%
-42.7 pp
39.8%
Python
82.0%
-42.2 pp
4.4%
Agile
44.0%
-39.6 pp
2.7%
AWS
40.0%
-37.3 pp
0.9%
Git
34.0%
-33.1 pp
1.8%
Azure
34.0%
-32.2 pp
0.9%
Docker
32.0%
-31.1 pp
0.9%
Kubernetes
26.0%
-25.1 pp
Company-pool skills derived from validated postings across Indian job boards for the data science & ml profile, scoped to two company classes: MAANG and Tier-1 Global Tech, and IT Services & BPO.