>_ AI / ML ENGINEER · CHENNAI, IN

Turning data
into decisions.

B.Tech – CSE (AI & Data Science) at Vel Tech University. I build practical AI systems, computer vision models, and data analytics pipelines — transforming data into measurable business outcomes.

View my work > Let's talk ↗
"I don't just write code — I build things that work in the real world."
Aakash S S - AI & Machine Learning Engineer
● MACHINE LEARNING
● DEEP LEARNING
● COMPUTER VISION
● NLP & RAG
● MLOPS ARCHITECTURE
● TENSORFLOW & KERAS
● PYTHON & FASTAPI
● EMBEDDED AI & IOT
● POWER BI & DAX
01 / PROFILE

The engineer
behind the code.

AI & Machine Learning Engineer — I build predictive models and embedded AI systems, from raw data to deployed hardware. My work spans computer vision, NLP, and data analytics, backed by hands-on experience across the full ML lifecycle: preprocessing, feature engineering, model evaluation, and stakeholder-facing reporting, gained through a data science internship at RINEX Technologies and an applied AI/embedded systems role at AICTE.

✦ Engineer by design. Problem-solver by nature.
01

Predictive Modeling & ML

Built and evaluated classification, regression, and NLP models — including a 96%-accuracy CNN classifier and a spam detector with 97% precision — using Scikit-learn, TensorFlow, and Keras.

02

Data Science & Analytics

End-to-end data preprocessing, EDA, and feature engineering on real-world datasets, translated into decision-ready dashboards using Power BI and DAX.

03

Embedded AI & Prototyping

Hands-on integration of AI models with ESP32-CAM and Arduino hardware at AICTE — rapid prototyping, testing, and deployment of AI-enabled embedded systems.

02 / SELECTED WORK

Proof, not
promises.

A curated selection of production-ready AI systems, predictive pipelines, and applied research projects built with measurable real-world outcomes.

01 MLOPS & PRODUCTION ML

Telco Customer Churn MLOps Platform

THE PROBLEM

ML models often work in notebooks but break down in production due to fragmented data pipelines, inconsistent environments, manual deployments, weak experiment lineage, and undetected model drift. This project builds an end-to-end, reproducible MLOps system that automates the journey from versioned customer data to monitored production inference and model promotion.

OUTCOME
0.847 ROC-AUC · Recall improved 48% → 80% through threshold optimization · Automated CI/CD, experiment tracking, deployment, monitoring & drift-aware retraining
Python · XGBoost · FastAPI · Docker · Kubernetes · MLflow · Prometheus/Grafana · Jenkins CI/CD · DVC
GitHub Repository ↗
02 CONVERSATIONAL AI & RAG

TN Government AI Scheme Assistant

THE PROBLEM

Government welfare information is fragmented across 37+ scheme documents and PDFs, making it difficult for citizens to quickly find relevant benefits, eligibility requirements, and application information. The challenge is to provide reliable English/Tamil answers from this scattered knowledge without hallucinating or returning irrelevant information.

OUTCOME
86/86 validation tests passed · 100% pass rate on 50-query benchmark · Bilingual RAG assistant deployed for real-time scheme discovery
FastAPI · Next.js · PostgreSQL · ChromaDB · Groq LLaMA 3.3 · RAG · Railway
GitHub Repository ↗ Live Demo ↗
03 COMPUTER VISION & EDGE IOT

AI-Based Smart Garbage Segregation Bin

THE PROBLEM

Manual waste segregation is inefficient, inconsistent, and difficult to scale. Disposal points need a low-cost autonomous system capable of identifying and separating multiple waste types in real time, while remaining functional when vision-based classification is unavailable.

OUTCOME
Real-time 5-category waste classification with sensor-based fallback, enabling autonomous segregation under both AI-enabled and degraded operating conditions.
Arduino (C++) · ESP32-CAM · Flask · TensorFlow
GitHub Repository ↗
04 MULTI-AGENT AI & PRODUCTION AUTOML

Autonomous Data Analyst (ADA)

THE PROBLEM

Enterprise data analysis is constrained by tedious manual workflows — diagnosing missingness, type coercion, masking sensitive PII, exploratory charting, baseline model tuning, compiling decks, and answering ad-hoc stakeholder questions without data hallucinations.

OUTCOME
7 specialized autonomous agents · 100% verified numerical grounding · Zero arbitrary code execution · Automated executive PDF/PPTX reports in < 25s · 590 unit tests passing (100%)
Python · Multi-Agent System · AutoML (LightGBM/XGBoost) · Streamlit · Flask REST API · n8n Automations
GitHub Repository ↗ Live Demo ↗

Looking for more projects, models, and open-source code?

Discover my complete collection of machine learning experiments, computer vision pipelines, NLP tools, and exploratory data analytics repositories on GitHub.

View All Projects on GitHub ↗
03 / TOOLKIT

Built across
the stack.

Languages
Python · SQL · Java (Basics)
AI & Machine Learning
Deep Learning · Computer Vision · NLP · Large Language Models (LLM) · Retrieval-Augmented Generation (RAG) · Scikit-learn · TensorFlow · Keras · XGBoost · CNN
Data Science & Analytics
Data Cleaning · Exploratory Data Analysis (EDA) · Predictive Modeling and Analytics · Statistical Analysis · Model Evaluation
MLOps & Deployment
Docker · Kubernetes · FastAPI · MLflow · CI/CD (Jenkins) · Prometheus/Grafana
Data Visualization
Power BI · Matplotlib · Seaborn · DAX
Embedded AI & IoT
Arduino UNO · ESP32-CAM · Internet of Things (IoT) · Sensor Integration · Embedded Systems
Tools & Practices
GitHub · VS Code · Jupyter Notebook · Google Colab

CERTIFICATIONS

✓ Infosys Springboard — Exploring Data Visualization · Feb 2026
↗
✓ RINEX (Google for Education Partner) — Artificial Intelligence · Nov 2025
↗
✓ RapidMiner (Altair) — Data Engineering Master Certification · Nov 2025
↗
✓ Deloitte — Data Analytics Job Simulation · Sep 2025
↗
✓ Tata Group — Data Visualisation: Empowering Business with Effective Insights · Sep 2025
↗
✓ Infosys Springboard — AWS Solutions Architect: Security & Data Privacy · Apr 2025
↗
✓ NVIDIA — Introduction to Networking · Apr 2025
↗
04 / EXPERIENCE & ACADEMICS

Background &
credentials.

Industry Experience

MAY 2026 – JUL 2026

AI & Embedded Systems Intern

AICTE, Chennai

Hardware & Software AI Engineering: Built a 5-category waste classifier (97.76% validation accuracy) via MobileNetV2 transfer learning with ESP32-CAM and moisture sensor fallback; engineered a cloud predictive analytics platform (Next.js, Supabase, Vercel) tracking child malnutrition risk and welfare delivery.

SEP 2025 – NOV 2025

Data Scientist Intern

RINEX Technologies, Bangalore

Worked on end-to-end Data Science and ML projects — data preprocessing, EDA, feature engineering, predictive modeling, and model evaluation on real-world business problems. Developed analytical reports and visualizations to communicate insights to stakeholders.

Education & Background

AUG 2023 – JUN 2027 · CURRENT

B.Tech – CSE (AI & Data Science)

Vel Tech University, Chennai
CGPA: 7.97
2023

All India Senior School Certificate Examination (AISSCE), CBSE

Jain Vidyaashram, Chennai
2021

All India Secondary School Examination (AISSE), CBSE

Jain Vidyaashram, Chennai
05 / DIRECT ANSWERS & FAQ

Clear answers for
teams & algorithms.

Structured facts and answers for recruiters, engineering managers, and AI answer engines seeking quick information about Aakash's technical competencies.

What makes Aakash different from other AI/ML engineering candidates?

+

Unlike most student ML portfolios built around notebook experiments, Aakash's work spans full production deployment — CI/CD pipelines, drift-aware retraining, containerized inference, and live-deployed applications — not just model training. His projects are shipped systems with monitoring, versioning, and real user-facing deployments, not one-off scripts.

What notable projects has Aakash built and what are the metrics?

+

Key benchmarked projects include:

• Telco Customer Churn MLOps Platform: 0.847 ROC-AUC, recall improved from 48% to 80% through threshold optimization, with automated CI/CD and drift monitoring.

• CropSmart-ML: Stacked ensemble (Random Forest + Optuna-tuned XGBoost + LightGBM + Extra Trees + SVM, with a Logistic Regression meta-learner) achieving 99.55% test accuracy on crop recommendation, outperforming four published benchmark papers. Available on GitHub ↗

• Autonomous Data Analyst (ADA): A 7-agent autonomous system that turns raw CSV/Excel files into verified business insights, automated PDF/PPTX reports, and a conversational Q&A interface — with zero arbitrary code execution, 100% numerically-grounded outputs, and a 590-test suite passing at 100%. GitHub Repository ↗ · Live Demo ↗

Does Aakash have real production/deployment experience, or just notebook-level ML?

+

Yes — the Telco Churn platform runs automated CI/CD, MLflow experiment tracking, Docker/Kubernetes deployment, and drift-aware retraining; the TN Government Scheme Assistant is a live, publicly deployed RAG application, not a local demo.

What industry and lab experience does Aakash have?

+

Aakash completed an AI & Embedded Systems Internship at AICTE (May 2026 – Jul 2026) integrating sensors, microcontrollers, and AI models into intelligent hardware. He also completed a Data Scientist Internship at RINEX Technologies, Bangalore (Sep 2025 – Nov 2025) engineering end-to-end data preprocessing, predictive modeling pipelines, and stakeholder analytics dashboards.

Is Aakash available for full-time roles, and is he open to remote or relocation?

+

Yes — Aakash is actively open to full-time AI/ML Engineering and Data Science roles, both remote and on-site. He's willing to relocate anywhere in India, with a preference for Chennai, followed by Bengaluru and Thiruvananthapuram. Reach out via email at aakash1552005@gmail.com.

06 / MAKE CONTACT

Have a problem
worth solving?

Whether it's a project or an opportunity — I'd like to hear from you.

+91 88259 09003
aakash1552005@gmail.com
→ Open in email client
LinkedIn
GitHub