Hello, I'm Amit Dutta, a Ph.D. candidate in Computer Science at the University of Nevada, Reno, focusing on adaptive learning systems. I have 6+ years of combined software engineering and applied AI research experience, building scalable ML pipelines, adaptive training systems, and full-stack applications using Python, PyTorch, LangChain, LangGraph, Spring Boot, React, and React Native.
My work includes peer-reviewed research on statistically effective adaptive training algorithms and production applications used by 10,000+ users.
Research, ML, and production applications that represent my work.
ResearchPhD research platform that turns a textbook into an adaptive Python course: RAG question generation against a topic taxonomy, deterministic sandboxed evaluation plus an LLM-as-judge layer aligned to professor feedback with GEPA, and Bayesian Knowledge Tracing that picks each student's next question from their weak topics.
ResearchNAT is a domain-agnostic, multi-parameter adaptive training algorithm that selects each learner's next activity from performance, weaknesses, and cognitive load; NAT-LLM generates after-action feedback from expert-written examples. Applied in a Unity-based naval simulator: NAT statistically outperformed non-adaptive training, and NAT-LLM reached 98% agreement with expert evaluations.
A stateful AI thinking-partner agent that refines a plan as you chat: one conversational agent with narrowly scoped tools where tool selection is the router, and deterministic precondition gates (proposal → authorization → commitment) instead of a graph of agents. Persistent memory in SQLite, streaming API, Next.js workspace.
A local, on-device AI dictation app for Windows: hold a hotkey, speak, release, and the transcription is inserted into the focused app. Local Whisper / Parakeet / Moonshine speech recognition with Silero VAD endpointing, local-LLM cleanup via Ollama with per-app style context, custom vocabulary and voice snippets, and a fine-tuned speech-correction model — no account, cloud, or telemetry.
Distilled a ViLT VQA model into a ~half-size student (6 layers, hidden size 384) under four losses — logit, hidden-state, attention, and hybrid — trained and evaluated through one shared harness on VQA v1.
End-to-end MLOps pipeline for anomaly detection on traffic sensor data, using MLflow and DVC for experiment tracking and reproducibility, model serving containerized with FastAPI and Docker, and Prometheus/Grafana monitoring for real-time metrics and observability.
Designed and evaluated a recurrent reinforcement learning approach that dynamically forms blocks in a Bitcoin blockchain network, balancing waiting time against block utilization. Published in IEEE Access.
BSc thesis: a PPO + RNN-LSTM trading agent in a custom OpenAI Gym environment, tuned via a 50-set technical-indicator search, that outperformed manual trading on the same indicators.
Trained a multi-head SEResNeXt50 classifier in PyTorch to jointly predict grapheme root, vowel, and consonant diacritics for Bengali handwriting. Submitted to a Kaggle competition.
Used KNN to build a simple food recommendation engine based on user preferences.
ResearchImplemented a rule-based system that adjusts scenario difficulty based on student performance in maritime navigation training.
MobileEnd-to-end white-label bike-sharing platform serving 10,000+ users. Mobile app in React Native with Firebase Authentication, Stripe payments, and CircleCI/CodePush deployment pipelines; backend fleet-management services in Spring Boot microservices with Google Cloud Pub/Sub event-driven ingestion for real-time device data across regions.
MobileA full-featured ERP system for managing HR, inventory, and sales. Built with React Native and Redux, available on Android and iOS.
WebResponsive website for Upay built using Next.js and Tailwind. Showcases services, charges, and features based on Figma designs.
A focused stack for building adaptive AI systems, ML pipelines, and production-ready applications.
Adaptive systems, model training, and reinforcement learning.
RAG pipelines, agentic workflows, and LLM evaluation.
Experiment tracking, reproducibility, and deployment monitoring.
Modern interfaces for web and cross-platform mobile apps.
APIs, services, databases, and cloud/payment integrations.
Game-engine simulations and native on-device applications.
University of Nevada, Reno
Pursuing a PhD in adaptive learning systems focused on combining machine learning and evolutionary computing. Developed neural network–based algorithms for personalized training.
2023 – PresentBrainstation23, Dhaka
Led development of Sharebike, a white-label bike-sharing app using React Native, Firebase, Stripe, and CodePush. Implemented Stripe for in-app payments. Built web apps with React and managed CI/CD pipelines using Fastlane and CircleCI. Also developed and maintained a microservice-based backend using Java Spring Boot. Mentored junior developers and created internal libraries to streamline development.
2021 – 2023Sweetitech, Dhaka
Developed multiple enterprise-level mobile and web applications using React, React Native, and Java Spring Boot. Built RESTful APIs, implemented MySQL database schemas, and integrated features like authentication, push notifications, and role-based access control. Contributed to ERP, HR, and pharmacy management systems deployed on both Android and iOS platforms.
2020 – 2021BRAC University Computer Club
Developed club management software using JavaFX and created design content using Photoshop and After Effects.
2018 – 2019NokshaIA, Dhaka
Designed and developed 2D games, implemented core gameplay mechanics, and contributed to game logic and UI programming.
2017 – 2018Peer-reviewed work connecting adaptive training, reinforcement learning, and applied AI systems.
Amit Dutta and Sushil J. Louis
Adaptive Instructional Systems, 8th International Conference, AIS 2026, 2026. Held as part of HCII 2026, Montreal, QC, Canada, July 26-31, 2026. Springer Nature, p. 301.
Amit Dutta and Sushil J. Louis
International Conference on Human-Computer Interaction, 2025. Springer, pp. 19-34.
Amit Dutta, Nafiz Imtiaz Rafin, M. Ali Akber Dewan, and Md. Golam Rabiul Alam
IEEE Access, 2024. Vol. 12, pp. 31287-31311. IEEE.
Biplob Kumar Daas and Amit Dutta
Journal of Materials Research, 2014. Vol. 29, no. 21, pp. 2485-2490. Springer.
Please contact me directly at [email protected] or through this form.