PYTHON • AI • ML • LLM • GENAI • DATA ANALYTICS

Applied AI & Python Development Internship

A project-driven programme that takes students from Python foundations to data analysis, machine learning, LLM integration and interactive AI applications — ending with an industry-style Banking Fraud Detection & Risk Explanation Dashboard.

Python NumPy Pandas Matplotlib Plotly Scikit-Learn LLMs Gemini PyTorch Huggingface Prompt Engineering

2Weeks

10Days

15hGuided Learning

15hProject Lab

Limited Seats ! Explore Curriculum

Internship Registration

* Required fields

Select your college

College *

Only colleges with a configured internship are searchable.

Name *

Gender *

Phone *

Email (Gmail Only) *

Programme / Degree *

Branch / Specialization *

Year *

Referral Code (optional) Referral code is optional. If entered, it will be checked for this session. Got a referral code? Add it here to receive credit.

Invite classmates. Get Rewarded.Top referrers across the programme will receive MeritNext Rewards. Referral credit is counted after successful payment.

By continuing, you agree to receive programme and payment-related communication from MeritNext. Registration is confirmed only after successful payment.

LEARN → BUILD → INTEGRATE → SHOWCASE

30 hours designed around doing, not just listening.

Every concept moves into code, visualization or an AI application. Students progressively build the components used in the final capstone.

Python Foundations

Syntax, data structures, functions, modules, exceptions, file/JSON handling and reusable code.

▥

Data + Visualization

NumPy, Pandas, Matplotlib and Plotly for cleaning, exploring and communicating transaction data.

ML / AI Algorithms

Classification, trees, ensembles, anomaly detection, model evaluation and risk thresholds.

✦

LLM + AI UI

Prompt engineering, structured output and Gradio interfaces that explain and visualize AI decisions.

2-WEEK CURRICULUM

From Python code to an explainable AI application.

A compact syllabus with enough depth to understand the technology stack and enough project time to use it.

WEEK 01

Python + Data + Machine Learning

Days 1–5

01Python FoundationsVariables, data types, operators, input/output, strings, lists, tuples, dictionaries, sets

02Programming Logicif/elif/else, for/while, functions, parameters, return values, scope, exceptions

03Python Packages & DataNumPy arrays, Pandas Series/DataFrames, CSV/JSON, filtering, grouping, missing values

04Data VisualizationMatplotlib & Plotly: distributions, trends, correlations, category comparisons, risk charts

05Machine Learning FundamentalsFeatures/labels, train-test split, preprocessing, classification, anomaly detection, evaluation

WEEK 02

LLMs + UI + Applied AI Project

Days 6–10

06ML Algorithms in PracticeLogistic Regression, KNN, Decision Tree, Random Forest, Isolation Forest + model comparison

07LLMs & Prompt EngineeringTokens, context, roles, zero/few-shot prompting, structured JSON output, hallucinations

08Mini AI ApplicationsAI Learning Assistant + AI Resume Analyzer using reusable prompt and Python patterns

09Gradio UI + IntegrationInputs, outputs, layouts, events, charts, model calls, validation and error handling

10Capstone Build & ShowcaseFraud scoring, explanation layer, dashboard integration, testing, documentation and demo

TECH STACK

Tools students will actually use.

The curriculum is intentionally application-focused: each library exists because it helps students build, analyze, model or present something.

Python Core

Variables • Collections • Loops • Functions • Modules • Exceptions • OOP basics • File handling • JSON

NumPy

ndarray • shape • slicing • vectorized operations • numerical features • aggregation

Pandas

DataFrame • read_csv • filtering • groupby • merge • missing values • feature preparation

VIZ

Visualization

Matplotlib • Plotly • histograms • bar charts • line charts • scatter plots • dashboards

scikit-learn

preprocessing • train_test_split • models • predict_proba • metrics • feature importance

LLM + Gradio

Prompt templates • structured output • API integration • interactive forms • data + AI UI

ML / AI ALGORITHMS

Students learn why different algorithms behave differently.

The goal is not memorizing formulas — it is learning how to choose, train, evaluate and explain models for a real problem.

Logistic Regression

Probability-based binary classification and a clean baseline for fraud prediction.

Supervised • Classification

K-Nearest Neighbors

Distance-based learning that makes feature scaling and similarity intuitive.

Supervised • Classification

Decision Tree

Human-readable rules, splits, impurity and overfitting — ideal for explaining decisions.

Supervised • Tree Model

♣

Random Forest

Ensemble learning, feature importance and stronger generalization for tabular risk data.

Supervised • Ensemble

◎

Isolation Forest

Detect unusual transactions without requiring every anomaly to be pre-labelled.

Unsupervised • Anomaly Detection

Kμ

K-Means

Clustering as a way to discover behavioural groups and compare transactions to peers.

Unsupervised • Clustering

Model EvaluationConfusion MatrixPrecisionRecallF1 ScoreROC-AUCThreshold TuningClass Imbalance

MINI PROJECTS

Build confidence before the final capstone.

Two smaller GenAI applications teach prompt design, Python integration and interface development before students combine the full stack.

MINI PROJECT 01

AI Learning Assistant

Generate explanations, key points, MCQs, answers and revision notes from a topic, learner level and difficulty.

Topic→Prompt→LLM→Structured Lesson→Gradio UI

Skills: functions • prompt templates • JSON • LLM API • UI inputs/outputs

MINI PROJECT 02

AI Resume Analyzer

Analyze a student resume against a target role and generate strengths, missing skills, interview questions and a learning roadmap.

Resume→Role→LLM→Skill Gap→Action Plan

Skills: text processing • prompt constraints • structured response • responsible AI

FINAL CAPSTONE

AI-Powered Banking Fraud Detection & Risk Explanation Dashboard

Students combine Python, data processing, visualization, machine learning, anomaly detection, LLM explanations and an interactive UI into one portfolio-ready application.

01Transaction DataCSV / DataFrame

→

02Feature EngineeringAmount • device • time • location

→

03ML Risk ModelClassification + anomaly score

→

04LLM ExplanationHuman-readable reasoning

→

05Gradio DashboardRisk + charts + action

Responsible AISynthetic/anonymized data, no real customer data

ExplainabilityModel scores are separated from LLM-generated narrative

Portfolio OutputWorking app + GitHub repo + README + final demo

Transaction MonitorIllustrative project UI

● MODEL ONLINE

TOTAL TXNS1,248

FLAGGED38

HIGH RISK12

IDAmountSignalRisk

#TX9821₹84,500New deviceHigh

#TX9822₹2,180NormalLow

#TX9823₹41,000Geo shiftMedium

#TX9824₹96,300Odd hourHigh

INTERNSHIP OUTCOMES

Students finish with demonstrable skills.

✓

Write structured Python programs and work with common data packages.

✓

Clean, explore and visualize data using Pandas, Matplotlib and Plotly.

✓

Train and evaluate supervised and anomaly-detection models.

✓

Integrate LLMs using structured prompts and controlled outputs.

✓

Build interactive AI interfaces using Gradio.

✓

Present a complete AI application with documentation and project demo.

CAPSTONE EXPERIENCE

See how Python, ML, LLMs and UI come together.

The final capstone turns transaction data into an explainable AI workflow — from feature engineering and fraud prediction to anomaly detection, visual analytics and an LLM-generated risk explanation.

DATA LAYERPandas • NumPy • Feature Engineering

MODEL LAYERClassification • Random Forest • Isolation Forest

EXPLAINABILITYRisk Score • Feature Signals • LLM Explanation

INTERFACEGradio • Plotly • Interactive AI Dashboard

AI Fraud Risk ConsolePython + ML + LLM

RISK SCORE87High Risk

ML PROBABILITY0.91Random Forest

ANOMALYYesIsolation Forest

Transaction SignalsFeature view

Amount92

New device78

Geo shift68

Odd hour54

LLM ExplanationAnalyst summary

This transaction is high-risk because the amount is unusually large, it originates from a new device and the location differs from the customer’s normal pattern.

Recommended: Step-up verification

★

PERFORM & GET RECOGNIZED

Perform well. Get Recognized. Get Awarded.

A Merit Test will evaluate learning from the internship. Top performers will receive a special prize and a Certificate of Merit from MeritNext.

⌘

LEARN & GROW TOGETHER

Learn together. Refer classmates. Get Rewarded.

Every confirmed student receives a personal referral link after payment. Share it with classmates — successful paid registrations are credited to the referrer, and top referrers receive special MeritNext Rewards.