Download Resume
M.Tech AI/ML Candidate • GenAI Engineer

Building applied AI systems that ship.

Aryan Kumar Sinha

I design and build NLP pipelines, RAG workflows, and LLM-powered applications with a strong focus on production-ready experimentation, retrieval, and real-world impact.

Get in Touch
Portrait of Aryan Kumar Sinha
Get to know me

About Me

About Aryan Kumar Sinha
4+ Years Coding

I am Aryan Kumar Sinha, an AI/ML postgraduate student based in Pune with hands-on experience across the applied GenAI stack.

My recent work spans smishing detection, geospatial place-change intelligence, EEG-to-image generation, and multimodal financial modeling. I enjoy turning messy real-world data into usable systems through structured experimentation, model evaluation, and strong product thinking.

I am especially interested in retrieval-augmented systems, agentic workflows, NLP, and production integrations that move beyond demos into reliable tools.

"From embeddings and vector search to prompt design and full-stack deployment, I like building AI products end to end."

— Current focus: applied GenAI engineering

Technical Skills

Python, Java, C, SQL
LLMs & Prompt Engineering
RAG, Agents & LangChain
NLP & Computer Vision
TensorFlow & Scikit-learn
FAISS, ChromaDB & Databases
Flask, Spring Boot & REST
OpenAI, Hugging Face & APIs
AWS, Docker, CI/CD & Linux
My work

Projects

Selected work across NLP, multimodal AI, geospatial intelligence, and automation, pulled from my recent resume and GitHub work.

SG Place Watch

Geospatial system for detecting new and closed businesses across 10 POI categories. It combines 15+ data sources, multilingual NLP, RAG validation, explainability reports, GeoJSON exports, and a Flask dashboard to process 3,500+ locations per run.

PythonRAGGeoJSON

DreamSketch EEG-to-Image

Generative AI pipeline that maps sleep EEG signals to semantic embeddings for image synthesis. It processes 2,000+ EEG-dream report pairs with bandpass filtering, STFT spectrograms, and occipital-channel extraction.

EEGGenerative AIPyTorch

Earnings Call Volatility Prediction

Multimodal pipeline combining earnings transcripts, audio sentiment, and historical stock data. It analyzes 2,000+ calls with Praat, FinBERT, wav2vec2, and ensemble models, achieving Spearman rho = 0.42.

FinBERTwav2vec2Finance ML

Smishing Detection System

Real-time smishing detection system using ANN models, TF-IDF, and word embeddings. The deployed mobile workflow reached 98% accuracy across 50,000+ SMS samples with sub-400ms prediction latency.

NLPANNCybersecurity

StoryScraper

Automation tool for downloading Instagram stories, reflecting my interest in practical scraping workflows, automation, and building utilities that save repetitive manual effort.

PythonAutomationScraping

YouTube Shorts Automation

Automation workflow for generating and publishing short-form video content, showing my comfort with external APIs, scripting repetitive media tasks, and production-minded automation.

PythonAPIsAutomation

Subway Surfer DDQN Agent

Deep reinforcement learning agent for autonomous Subway Surfer gameplay. Built with a DDQN architecture and a TFX training, evaluation, and inference pipeline, reaching a score of 14,569,684.

Deep RLDDQNTensorFlow

Cat vs Dog CNN Classifier

Early deep learning project focused on image classification with convolutional neural networks, marking the foundation of my longer-term move into AI and model-building.

CNNComputer VisionDeep Learning
View More on GitHub
Experience and learning

Experience & Credentials

The internship, education, and certifications that define my current AI/ML path.

Research Intern
La Trobe University
October 2023 – April 2024

Built a real-time smishing detection system with ANN models, NLP feature pipelines, and a REST-integrated mobile application for fast user-facing predictions.

Key Highlights:

  • 98% classification accuracy across 50,000+ SMS samples
  • Reduced threat analysis time by 30% and improved detection speed by 40%
  • Delivered sub-400ms predictions through a deployed mobile workflow
Click to expand
Master of Technology in AI & ML
Symbiosis Institute of Technology
August 2025 – May 2027

Currently pursuing postgraduate specialization in artificial intelligence and machine learning while continuing to build applied research and production-oriented AI projects.

Key Highlights:

  • Program focus on AI/ML methods, experimentation, and deployment
  • Current GPA: 7.15
  • Research interests include RAG, LLM systems, NLP, and multimodal AI
Click to expand
B.Tech in Computer Science & Engineering
Vel Tech Rangarajan Dr. Sagunthala R&D Institute
August 2020 – May 2024

Undergraduate foundation in computer science, software development, and machine learning that set up my transition into applied AI systems work.

Key Highlights:

  • CGPA: 8.57
  • Built the base for later work in ML, NLP, and software engineering
  • Strong grounding in programming, databases, and full-stack problem solving
Click to expand
Certifications
DeepLearning.AI, IBM, Databricks, NPTEL, Udemy
2023 – 2026

Continuous upskilling through certifications spanning LangGraph agents, Databricks data engineering, Kafka streams, generative AI engineering, and applied data analysis.

Key Highlights:

  • AI Agents in LangGraph by DeepLearning.AI x LangChain (2026)
  • Generative AI Engineering with LLMs Specialization by IBM x Coursera (2025)
  • Data Engineering using Databricks on AWS and Azure by Udemy (2025)
  • Apache Kafka Series: Kafka Streams for Data Processing by Udemy (2025)
  • Data Lake Mastery: The Key to Big Data & Data Engineering by Udemy (2025)
  • Data Analysis for Biologists by NPTEL (2023)
  • Architecture With VMware NSX by Coursera (2020)
Click to expand
98% Best Classification Accuracy
50k+ SMS Samples Analyzed
3.5k+ Locations Processed
8 Resume Certifications