Aswin N S
Machine Learning & Software Engineer
B.Tech in Artificial Intelligence & Data Science graduate from Karunya Institute of Technology and Sciences. Specializing in sequential deep learning (Bi-LSTM, attention mechanisms), predictive analytics, and production-ready APIs with FastAPI.
Attention-based Bi-LSTM architectures paired with SHAP (SHapley Additive exPlanations) for 24-hour migraine forecasting and water quality time-series prediction.
Built during industry internship at AkooBooks Audio Ltd. Converting unstructured natural language queries into dense embeddings for conceptual catalog discovery.
About Me
AI & Data Science graduate from Karunya Institute of Technology and Sciences, building practical predictive systems, explainable models, and RESTful API backends.
I recently graduated with a B.Tech in Artificial Intelligence and Data Science from Karunya Institute of Technology and Sciences, Coimbatore. I focus on practical machine learning, exploratory data analysis, and building reliable software backends that connect intelligent models to end users.
My primary technical interests are sequential deep learning (Bi-LSTMs and attention mechanisms), predictive time-series modeling, and explainable AI (XAI). For instance, in my PERCEPT-MIND migraine research project, I paired an attention-based Bi-LSTM with SHAP (SHapley Additive exPlanations) so that every prediction clearly highlights which physiological and environmental factors contributed to the forecast.
Beyond machine learning models in notebooks, I care about engineering practical applications. During my AI internship at AkooBooks Audio Ltd., I developed a semantic search pipeline for an AI-powered virtual librarian, converting natural-language user queries into embeddings and serving ranked results through a FastAPI backend. In earlier internships at Redintek and Gogaga Entertainment, I worked on REST APIs and structured ERP data processing.
I am based in Dubai, UAE, and actively looking for entry-level AI/ML, data science, data analyst, or software development opportunities in the UAE and internationally. I bring solid fundamentals in Python, SQL, machine learning, and API development, along with a strong desire to learn and contribute to real-world engineering teams.
Target Engineering Positions
B.Tech in AI & Data Science
Karunya Institute of Technology and Sciences, Coimbatore
Graduating with solid foundations in machine learning, deep neural networks, time-series analysis, DBMS, and software engineering.
AkooBooks Audio Ltd. AI Internship
AI & NLP Intern (Remote, South Africa)
Engineered semantic retrieval pipelines using text embeddings and FastAPI REST services to assist visually impaired and audio-first listeners.
Healthcare Research & SHAP
PERCEPT-MIND Framework
Developing attention-based Bi-LSTM sequential models paired with SHAP to deliver transparent risk attribution for preventative healthcare.
Located in Dubai, UAE. Available immediately for on-site, hybrid, or remote engineering positions across the UAE (Dubai, Abu Dhabi, Sharjah) as well as global opportunities.
Skills & Engineering Tooling
A comprehensive breakdown of tools, frameworks, and methodologies applied across academic research, industry internships, and production-ready applications.
Machine Learning & Deep Learning
Sequential Architectures & Predictive Modeling
Designing and validating deep learning models with a focus on temporal sequences, attention layers, and interpretability.
Neural network training, loss optimization, model evaluation
Temporal sequence modeling for 24h risk & water index
Scaled additive attention to weight influential time steps
Regression, classification, preprocessing, cross-validation
MAE, RMSE, R², ROC-AUC, confusion matrices
Tokenization, Named Entity Recognition in legal text
Backend & API Engineering
Asynchronous Services & RESTful Architectures
Connecting data pipelines and machine learning inference to high-throughput REST APIs and microservice architectures.
Asynchronous endpoints, Pydantic schemas, automated OpenAPI docs
REST routing, middleware pipelines (Redintek internship)
Clean status codes, structured JSON payloads, error handling
Containerizing Python services for reproducible environments
Decoupled frontend-backend interaction patterns
Data Analytics & Preprocessing
Feature Engineering & Exploratory Analysis
Transforming noisy environmental and physiological signals into normalized, structured datasets ready for model ingestion.
Outlier detection, null handling, MinMaxScaler normalization
Dataframe transformations, vector math, sliding window creation
Correlation heatmaps, distributions, statistical profiling
Data visualization and diagnostic loss curve plotting
Interactive geographic mapping of environmental well sites
Business KPI reporting and analytical dashboard layouts
Databases & Storage
Relational & Document Data Management
Designing relational schemas, writing optimized analytical queries, and managing semi-structured document payloads.
Foreign keys, relational constraints, indexing, ACID transactions
Relational table design, multi-table joins, administrative queries
Aggregations, window functions, ERP workflow data cleanup
Document store for semi-structured search payloads & logs
Programming Languages & Web
Core Foundation & Modern Interfaces
Strong algorithmic foundations with Python as the primary language, backed by modern full-stack web capabilities.
Used across all machine learning models, ETL, and FastAPI APIs
Analytical data extraction, database schema creation, table joins
Asynchronous event loops, REST client integration, Node.js
Static type checking, interfaces, and component prop validation
Building responsive, accessible, high-performance web UIs
Branching strategies, version control, code review collaboration
Explainable AI & Research
Interpretable Healthcare Decision Support
Pioneering model transparency in high-stakes predictive analytics where black-box decisions are unacceptable.
Global & local feature attributions for risk forecasting
Integrating physiological, behavioral, and weather signals
Preventative action trigger logic based on risk score thresholds
Local surrogate explainability for tabular & textual model outputs
Featured Engineering Projects
Real software systems and machine learning implementations spanning time-series deep learning, attention mechanisms, explainable AI (SHAP), semantic search, and RESTful backends.
PERCEPT-MIND: Migraine Prediction Research
Personalized Migraine Risk Management Using Attention-Based Deep Learning & Explainable AI
A proof-of-concept predictive analytics framework for personalized migraine management using attention-based deep learning and explainable AI.
Formulated a multi-stage sequential analytics pipeline combining attention-based Bi-LSTM networks to predict 24-hour migraine onset likelihood from multi-source signals, paired with SHAP interpretability to explain feature contributions transparently.
Water Quality Prediction System
Sequential Deep Learning for Environmental Parameter Modeling
A machine learning and deep learning project exploring water quality prediction across temporal sequences using an environmental dataset.
Constructed a multi-parameter machine learning pipeline that preprocesses water chemistry data, applies sequential sliding windows, and trains deep LSTM models to forecast overall water quality index metrics.
AI Virtual Librarian for Audiobook Search
NLP Semantic Retrieval Pipeline & Assistive Search Architecture
An assistive technology project designed to improve access to information and audiobook discovery for visually impaired and audio-first listeners.
Developed an NLP-driven semantic query pipeline that extracts query intent, generates contextual embeddings, and retrieves relevant audio content via a FastAPI backend service.
Smart FIR Automation System
NLP Information Extraction & Legal Text Processing Backend
An NLP-based project involving FIR-related text processing and backend application integration to extract structured details from complaint narratives.
Created an information extraction pipeline utilizing Named Entity Recognition (NER) and tokenization to extract key legal variables from complaint statements into structured database payloads.
Truth Checker — Misinformation Analysis Prototype
NLP Text Analysis for Assessing Potentially Misleading Information
An NLP-related project exploring text analysis and semantic consistency scoring to assess potentially misleading claims.
Engineered an experimental claim analysis pipeline that computes semantic similarity, lexical indicators, and model confidence scores to flag claims requiring manual verification.
Consulting Matchmaking Platform
Full-Stack Client-Expert Recommendation System
A modular web application matching business clients with specialized consultants based on domain requirements and skill parameters.
Architected a full-stack platform featuring a Python FastAPI backend for inquiry matching and a responsive React frontend for client-consultant interactions.
Experience & Education
Verified internship track record, degree coursework at Karunya Institute of Technology and Sciences, and industry credentials.
Practical Internships
AI Intern
Jun 2025 — Sep 2025- •Developed an AI-powered virtual librarian for audiobook search utilizing NLP and semantic retrieval techniques.
- •Built query preprocessing, embedding generation, and contextual similarity matching pipelines.
- •Integrated the NLP pipeline with a high-performance FastAPI backend and user interface for responsive search.
Full Stack Developer Intern
Jan 2024 — Mar 2024- •Developed backend APIs and supported structured data-processing workflows.
- •Automated recurring internal data workflows, improving communication between application services.
- •Earned 3rd Place in the Redintek Company Hackathon for collaborative engineering.
Software Intern
Aug 2023 — Nov 2023- •Processed and optimized structured ERP data workflows to support internal operational efficiency.
- •Assisted in data verification, record integrity checks, and internal reporting.
Academic Degree
- Comprehensive training in Machine Learning, Deep Learning, Natural Language Processing, Database Management Systems (DBMS), and Software Engineering.
- Active researcher in healthcare predictive analytics, developing attention-based BiLSTM models and explainable AI pipelines.
- Practical laboratory and project experience building machine learning models and web-integrated applications.
Certifications & Awards
Microsoft Certified: Azure AI Fundamentals (AI-900)
Microsoft
Python Programming
Cisco Networking Academy
Cybersecurity Essentials
Cisco Networking Academy
Foundations of Cybersecurity
3rd Place — Hackathon Winner
Redintek Company
Research & Publications
Investigating sequential attention architectures and explainable AI (SHAP) for transparent, preventative healthcare risk forecasting.
PERCEPT-MIND: A Predictive Analytics Framework for Personalized Migraine Management Using Attention-Based Deep Learning and Explainable AI
Proposes a multimodal predictive framework combining sequential physiological, environmental, and lifestyle factors with attention-based Bi-LSTM networks for 24-hour-ahead migraine onset risk forecasting, interpreted with SHAP feature explanations.
Core Methodological Contributions
Get in Touch
Send an inquiry directly to Aswin's personal inbox at aswinsn44@gmail.com with real-time delivery and fast response times.
Current Location
Dubai, UAE
Based in Dubai, UAE (Available for on-site & remote roles)