Exeter, United Kingdom · Open to data science and ML roles

ManuelAugustineFaria

Data Scientist, Machine Learning Engineer, AI Engineer, Computer Vision Engineer, AI Developer

I build predictive models, computer vision systems and multi-agent AI tools, with honest evaluation and stated limits.

Data
Analysis
Model
Decision

View projects Contact me

Projects · 13

Go to the section

Resume

Python
ML
AI
Computer Vision
GenAI

Hover or focus a stage

Scroll

01About

From mechanical engineering to data science.

I started in mechanical engineering (Diploma, 91%, Distinction), moved into computer science with a BTech in AI and Machine Learning, and completed an MSc in Data Science at the University of Exeter with Merit.

That route explains how I work: I like precise, well-measured systems, and I distrust a result I cannot explain. I have applied ML in client work and data engineering, and carried the same habits into research.

Research interests: computer vision, time-series forecasting, robustness and evaluation, and safety-critical or societal-impact domains.

Currently

Graduated: MSc Data Science with Merit, University of Exeter
Building: The live labs on this site, each rebuilt to run in the browser

Education

2025 – 2026
MSc Data Science
University of Exeter · Merit
2021 – 2024
BTech Computer Science & Engineering
Sri Ramachandra Institute of Higher Education and Research · CGPA 7.83 / 10 · First Class
2018 – 2021
Diploma in Mechanical Engineering
Panimalar Polytechnic College · 91% · First Class with Distinction

Hackathon results

1st prize, team result
Course recommendation
3rd prize, team result
AR navigation

Certificates

Cloud Foundations · AWS Academy Graduate
Introduction to Cloud · AWS Academy Graduate
Blockchain A-Z: Blockchain, Cryptos & Smart Contracts · Udemy
Machine Learning · Simplilearn

02Experience

Experience

Client work, data engineering, five earlier internships, and the part-time job alongside the MSc.

May 2026 – Present · Part-time
Retail Assistant
Primark
Customer-facing work alongside my studies.
- Customer-facing communication, operational discipline, working under pressure and teamwork in a fast-moving retail environment.
- Communication
- Teamwork
- Adaptability
Aug 2024 – Aug 2025
Freelance AI/ML & Software Consultant
Independent client work
Delivered end-to-end applied ML for clients, and built AI-powered websites with chatbot and predictive features.
- Built demand-forecasting and inventory-analysis workflows in Python on retail sales data; the analysis informed purchasing and stock decisions.
- Applied statistical analysis to surface KPI insights and presented them to non-technical stakeholders.
- Designed and built AI/ML-powered websites for clients, including chatbots.
- Python
- Forecasting
- Statistics
- Chatbots
- Client communication
May 2024 – Aug 2024
Data Engineering Intern
Agilisium Consulting
Worked on enterprise data cleaning, transformation and KPI reporting.
- Developed a Power BI dashboard covering 511 employees.
- Supported data cleaning, transformation and pipeline documentation for reporting workflows.
- Power BI
- Data cleaning
- ETL
- KPI reporting
Mar – May 2024
AI and 3D Simulation Developer Intern
Sri Ramachandra
Project: Driver behaviour simulation
Jan – Mar 2024
Machine Learning Intern (network security)
Sri Ramachandra
Project: DDoS detection
Dec 2022 – Feb 2023
AR Development Intern
Sri Ramachandra
Dec 2022 – Feb 2023
Front End Developer
HEDITH
My first React website (opens in a new tab)
Dec 2021 – Feb 2022
Time Series Analysis Intern
Sri Ramachandra
Project: Time-series forecasting

03Projects

Selected work

The dissertation and three projects that show the range best: the problem, what I did, and what it was built with. Each opens the full case study.

All 13 projects

Research

2026 · Generative AI

Human–Multi-Agent Teaming for Higher-Order Thinking

Problem: Single-agent chat tutors tend to hand students finished answers, which flattens the higher-order thinking that data science education is meant to build.
Approach: Interaction telemetry was logged per session and analysed turn by turn, classifying student turns and tracking how the share of higher-order-thinking interactions changes as a session progresses.
Built with: Python · LangGraph · FastAPI · SHAP · LIME · LLMs

5.6% → 16.5%

Higher-order-thinking interactions, turn 1 → 6

Coursework

2026 · Computer vision

Campus Safety Object Detection

Problem: Object detection is easy to demo and hard to trust. Models that look accurate on curated test sets fail on occlusion, small objects and edge-case lighting, which matter most in safety-critical settings.
Approach: A YOLOv8n detector was trained on public COCO images combined with campus-specific photographs, and evaluated with mAP, IoU and confidence-threshold analysis, including targeted occlusion and small-object test cases.
Built with: Python · YOLOv8 · OpenCV · PyTorch · Computer Vision

5 classes

Person, bicycle, car, stop-sign, traffic-light

Live lab
Internship
Synthetic data

2024 · Machine Learning

DDoS Attack Detection using Supervised Learning

Problem: Software-defined networks concentrate control logic, which makes them efficient and a high-value DDoS target. A detector that is accurate but raises too many false alarms is operationally unusable.
Approach: Network-traffic features were extracted and fed through multiple candidate classifiers, with a feature-selection stage used to compare which signals carry detection value versus noise. Because attack traffic is rare, accuracy alone was not trusted as a metric.
Built with: Python · scikit-learn · Pandas

P ↔ R

Precision–recall trade-off, threshold tuned

Open the live lab

Live lab
Hackathon
Team result
Synthetic data

2024 · Recommendation systems

Course Recommendation System

Problem: A course catalogue can hold many options, and recommendations can help people find the ones that are relevant to them. The project looked at how a recommendation-based approach could do that.
Approach: The live prototype above is new work built for this portfolio. It is not the hackathon submission. A learner picks an area, a target topic, their level, what they already know, a preferred format and any courses already completed. Hand-written rules remove courses already completed or marked not interesting. A decision tree then classifies each remaining course as a good fit, a stretch or not suitable, a hand-weighted score orders the courses within each class, and each result shows its reasons, its prerequisites and the path the profile took. A learning map shows the topics of the area in prerequisite order.

1st prize

Team result · Agilisium Consulting hackathon

1st Prize · Hackathon · team result

Open the live lab

04All projects

Every project

13 projects, newest first: research, coursework, internships, hackathons and personal builds. Open one for the full story; the live labs run in your browser.

- Featured
- Research
Human–Multi-Agent Teaming for Higher-Order Thinking
Designed and evaluated a nine-agent LLM learning environment for AI and Data Science education, analysing interaction telemetry to study higher-order thinking behaviour.
Built with: Python · LangGraph · FastAPI · SHAP
2026 · Generative AI
- Featured
- Coursework
Campus Safety Object Detection
Built a YOLOv8n object detector for person, bicycle, car, stop-sign and traffic-light classes using public COCO images and campus-specific photographs.
Built with: Python · YOLOv8 · OpenCV · PyTorch
2026 · Computer vision
- Coursework
Sleep Duration vs Academic Performance
A statistical study of sleep duration and academic performance among university students, with a deliberate focus on the limits of causal inference.
Built with: Python · Pandas · Regression · Statistical analysis
2025 · Statistics
- Featured
- Live lab
- Hackathon
- Team result
- Synthetic data
Course Recommendation System
A team project built during my Data Engineering internship at Agilisium Consulting, exploring a recommendation-based approach to finding relevant courses. The team won first prize. This page also hosts a new explainable course-recommendation prototype covering 13 domains.
2024 · Recommendation systems
- Live lab
- Internship
- Synthetic data
AI-Based Driver Behaviour Simulation
Modelling how drivers respond when the road stops behaving predictably. An Unreal Engine 5 internship prototype, rebuilt as a 2-D top-view simulation: pick a start and a destination on the map and an AI driver follows the road rules there.
Built with: Blender · Unreal Engine 5 · Python · scikit-learn
2024 · Simulation
- Featured
- Live lab
- Internship
- Synthetic data
DDoS Attack Detection using Supervised Learning
Trained supervised classifiers to distinguish malicious from legitimate network traffic in a heavily imbalanced dataset.
Built with: Python · scikit-learn · Pandas
2024 · Machine Learning
- Live lab
Smart Attendance System
An automated computer-vision attendance system tested across lighting and pose variation, with biometric privacy and responsible-AI considerations built into the evaluation.
Built with: Python · OpenCV · Deep learning · Face recognition
2023 · Computer vision
- Live lab
- Coursework
Fake Product Detection using Blockchain
A product-authentication platform that links QR codes to tamper-evident blockchain records, so a customer can check that a product identity matches what the manufacturer registered. Try the live demo.
Built with: Solidity · Ganache · MetaMask · QR codes
2023 · Blockchain
My First Portfolio Website
The first portfolio I built, in July 2023, with React.js. It is live at manuelfaria2702.netlify.app.
Built with: React.js
2023 · Web development
- Live lab
Explainable Machine Learning for Performance Prediction
Built and compared Logistic Regression, Random Forest and XGBoost classifiers using preprocessing, exploratory analysis, feature engineering and k-fold cross-validation.
Built with: Python · scikit-learn · XGBoost · SHAP
2023 · Machine Learning
- Live lab
- Internship
Time-Series Forecasting of Financial Data
Stationarity testing, ACF/PACF analysis and AR, MA, ARMA, ARIMA and SARIMAX forecasting on historical datasets, including Air Passengers and Dogecoin. Play the live demo to watch the model forecast step by step.
Built with: Python · statsmodels · ARIMA · SARIMAX
2021–2022 · Time series
Gesture-Controlled Wheelchair for Paralyzed People
A Diploma engineering project exploring a low-cost gesture-controlled wheelchair prototype using MEMS sensing, Arduino microcontrollers, RF wireless communication, relay control and DC motors.
Built with: Arduino · RF · MEMS sensor · Relays
2021 · Mechanical engineering
- Hackathon
- Team result
AR Navigation Prototype
An augmented-reality navigation idea for a university setting, built into a small prototype to show the concept. The team received third prize in a hackathon.
Undated · Augmented reality

05Skills

The tools behind the systems I build.

Grouped by what I use them for. The case studies show each one in use.

Programming: Python · SQL · R
Data science and ML: Pandas · NumPy · scikit-learn · XGBoost · Random Forest · TensorFlow · Keras · PyTorch
Vision and AI: OpenCV · YOLOv8 · LLM applications · Retrieval-augmented generation · Multi-agent systems
Analysis and visualisation: Statistical analysis · Matplotlib · Power BI · Tableau
Web development: React.js · REST APIs
Tools and platforms: Git · GitHub · Docker · AWS · Linux

06Contact

Let's build something useful.

Open to opportunities across Data Science, Machine Learning, Applied AI, Computer Vision, Data Engineering and research.

Exeter, United Kingdom

Send a message

ManuelAugustineFaria

From mechanical engineering to data science.

Currently

Education

Hackathon results

Certificates

Experience

Retail Assistant

Freelance AI/ML & Software Consultant

Data Engineering Intern

AI and 3D Simulation Developer Intern

Machine Learning Intern (network security)

AR Development Intern

Front End Developer

Time Series Analysis Intern

Selected work

Every project

The tools behind the systems I build.

Let's build something useful.