Inesh Tandon

Software Engineering • Data Science • Machine Learning

University of Maryland, College Park, MD

PythonJavascriptDockerAWSPyTorchR

Building problem solving solutions

I am passionate about designing and developing scalable AI systems, extract data-driven insights, and innovative software solutions that make a meaningful impact.

Projects

FlaskLangChainWeaviateDocker
Generative AI integrated Extension
Personal AI Agent for everyday tasks
  • Built a Chrome extension with native UI integration to capture email context, tone, and intent, generating tailored responses via a RAG pipeline using LangChain, Dockerized Flask and Weaviate
  • Developed a dynamic frontend using JavaScript-injected HTML/CSS for near real-time email generation, improving user workflow efficiency
60% reduced email drafting timeRAG Optimized User Personalization
LangChainFastAPIFAISSATS Optimization
RAG Guided Resume Generator
AI-Powered Job Application Assistant
  • Built a FastAPI-hosted LangChain RAG system with advanced prompt engineering for ATS-optimized resume bullet points
  • Integrated FAISS-powered retrieval to augment resumes with targeted guidance for job-specific customization
LLM-Enhanced GenerationTailored Resume Optimization
PyTorchTemporal CNNTime-SeriesAttention
Multi-Time Series Risk Prediction
Financial Volatility Forecasting
  • Developed a 2-tower neural network with attention-based Temporal CNN and 16 engineered temporal features across 20 years of data
  • Performed 3-class classification for 23 datasets over a 15-day horizon, achieving 14% higher F1 score over LSTM baselines using hybrid normalization
14% ↑ F1 Score20 years Data
Reinforcement LearningSoft Actor-CriticGymnasiumEnergy
Smart HVAC Controller
Energy Optimization with RL
  • Designed a multi-objective HVAC controller with Soft Actor-Critic, balancing comfort and efficiency under real-world constraints
  • Simulated temperature, humidity, and wind in a custom Gymnasium environment to model user preferences
81% Energy Savings3-Member Team
PyTorchGANAlexNetComputer Vision
Image Inpainting with GANs
Context Encoders for Image Reconstruction
  • Built a masked dataset of 10K+ VOC2012 images with structured dropout for training
  • Trained AlexNet-inspired context encoder with masked L2 and adversarial loss, achieving realistic image inpainting
  • Shared technical insights in a Medium article on model design and training challenges
10K+ Training ImagesPublished Medium Article

Experience

Feb 2025 – Present • DoQuantum UMD
Quantum ML Researcher
  • Purpose: Explore quantum-classical hybrid methods to enhance transformer attention efficiency and scalability.
  • Work: Designed a transformer attention module that delegates the query–key similarity computation to a quantum routine; integrated the subroutine into a PyTorch pipeline for end-to-end experimentation.
  • Impact: Validated feasibility of quantum-assisted attention on real quantum hardware, providing benchmarks for potential speedup and accuracy improvements.
Hybrid Attention Model Quantum Hardware Validation PyTorch Integrated
Apr 2024 – Jun 2024 • MindShift Analytics
Data Scientist
  • Purpose: Ensure reliability and accuracy of predictive reporting at scale.
  • Work: Designed automated Python + MySQL pipelines with embedded sanity checks; built and iteratively improved backend validation for 1K+ predictive reports.
  • Impact: Eliminated update failures and boosted report accuracy through robust validation.
100% ↓ update failures 35% ↑ report accuracy 1K+ reports validated
Apr 2023 – Jun 2023 • Lok Smriti Sewa Sansthan
Data Manager
  • Purpose: Improve medicine stock availability and patient retention.
  • Work: Built regression models on inventory and follow-up data; standardized data collection and reporting workflows.
  • Impact: Enabled data-driven restocking and outreach decisions that lifted key operational outcomes.
25% ↑ stock availability 20% ↑ patient retention

Education & Certifications

MS, Data Science — University of Maryland (Expected 05/2026)

Advanced ML • Advanced Algorithms • Big Data/Distributed Systems • Computer Vision • Natural Language Processing • Probability & Statistics

BE, Information Science & Engineering — VTU (07/2024)

Data Structures • Object Oriented Programming • Database Management • Software Development Lifecycle • Operating Systems

Contact

Email GitHub LinkedIn

Let's Build Something Amazing

Ready to collaborate on innovative projects? I'm always excited to discuss new opportunities in AI, data science, and software engineering.