Aspiring Software & ML Engineer

Mohammed Adnan Aziz

Computer Science Graduate · Aspiring Software & ML Engineer

+91-970******* *****.*********@*****.*** linkedin.com/in/mohammed-adnan-aziz-857030206 Hyderabad, India

OBJECTIVE

I'm a 2025 Computer Science graduate who genuinely enjoys building things that work — whether that's a machine learning model

that predicts flood risk in real time or a Flask API that tells you if a UPI payment will go through. I've had the chance to intern at IBM

SkillsBuild and Huntmetrics, and those experiences taught me that the best solutions come from understanding a problem deeply

before writing a single line of code. I'm looking for a role where I can keep learning, contribute meaningfully from day one, and grow

alongside a team that takes its craft seriously.

EDUCATION

B.E. in Computer Science Engineering 2021 – 2025

Lords Institute of Engineering & Technology, Hyderabad

Intermediate — MPC (Maths, Physics, Chemistry) 2019 – 2021

Narayana Junior College · Telangana Board of Intermediate Education

Secondary School Certificate 2019

Royal Embassy School

TECHNICAL SKILLS

Languages Python, Java, JavaScript, SQL

Web & APIs HTML5, CSS3, Flask

ML & Data Scikit-learn, XGBoost, Pandas, Matplotlib, Seaborn, SMOTE

Hardware/IoT Arduino, Raspberry Pi, Ultrasonic Sensors, Servo Motors

Other Tools Git, IBM SkillsBuild, VS Code

Soft Skills Team leadership, clear communication, deadline-driven, collaborative

INTERNSHIPS

AI & Machine Learning Intern — IBM SkillsBuild June – July 2024

Project: Real-Time Flood Detection & Prediction using ML

• Built a machine learning model from scratch to predict flood risk using historical weather records and live sensor inputs — the

kind of tool that can genuinely help communities prepare before disaster strikes.

• Cleaned and preprocessed large, messy datasets, which honestly taught me more about real-world data than any classroom

could.

• Trained and compared several ML algorithms, tuning each one to balance accuracy with real-time performance requirements.

• Integrated live data streams so the model could update risk assessments dynamically, not just on static snapshots.

Cyber Security Intern — Huntmetrics, India August – September 2023

Project: Identifying Insecure Deserialization Sources

• Investigated how insecure deserialization vulnerabilities (an OWASP Top 10 risk) can silently compromise application security —

and built methods to surface them before attackers do.

• Used security testing tools to probe serialization and deserialization flows in real applications, spotting weaknesses that code

reviews often miss.

• Documented and implemented mitigation strategies — input validation, sanitization — that left the codebases measurably

more secure.

PROJECTS

Predicting UPI Transaction Success & Failure

Python · Pandas · Scikit-learn · XGBoost · Flask · Matplotlib

• UPI failures are frustrating for everyone involved. I wanted to understand why they happen, so I collected transaction data with

features like bank, device, network type, time of day, and transaction amount — then built models to predict outcomes before

they occur.

• Applied SMOTE to deal with the natural class imbalance (failures are rarer than successes), and ran EDA that revealed time-of-

day and network strength as the biggest culprits.

• Compared three classifiers, settled on XGBoost for its interpretability and performance, then wrapped it in a Flask API so it could

serve real-time predictions.

• The end result: actionable insights into what drives failures, packaged in a form that a fintech team could actually use.

Car Insurance Claim Amount Prediction

Python · Pandas · Scikit-learn · XGBoost · Flask · Seaborn

• Insurance companies lose money when claim amounts are unpredictable. I built a regression pipeline to estimate claim severity

from policyholder profiles, vehicle data, and claim history.

• Engineered meaningful features like vehicle age and claim-to-premium ratio that turned out to be strong predictors the raw

data didn't make obvious.

• Ran hyperparameter tuning and cross-validation across multiple models, achieving a high R score with low MAE and RMSE —

the model generalises well, not just fits the training data.

• Deployed via Flask so underwriters could get real-time predictions and adjust premiums accordingly.

ACHIEVEMENTS

2nd Place — Talent Hunt 2K24 2024

Lords Institute of Engineering & Technology · Project: IoT-Based Radar System

• Built a working radar prototype using ultrasonic sensors wired to an Arduino and controlled by servo motors — the kind of

project where everything has to come together perfectly or nothing works.

• Visualised live object-tracking data on a custom GUI that mimicked real radar sweeps. Competing against cross-disciplinary

teams, we secured second place — which felt like a win given how strong the field was.

Contact this candidate