Built an Android AI agent that can see the phone screen, understand voice commands, and complete multi-step mobile tasks in real time. The cloud branch uses model APIs for fast autonomous screen control, while the on-device branch explores private local inference with speech-to-text, a vision-language model, and text-to-speech running directly on the phone.

Built a complete Speech-to-Text Transformer model from scratch using PyTorch, converting raw audio waveforms into text without pre-trained models. Implements convolutional downsampling, multi-head self-attention, Residual Vector Quantization (RVQ), and CTC loss for alignment-free training. Trained on the LJSpeech dataset using an A100 GPU.

Implemented a Tacotron 2 neural text-to-speech model from scratch in PyTorch. The model generates mel-spectrograms from raw text input using an encoder-decoder architecture with attention mechanisms, then converts them to audio waveforms. Trained on the LJSpeech dataset.

Built an OpenEnv-native scalable oversight gym for hidden sabotage detection, inspired by SHADE-Arena. A frozen DeepSeek-R1 attacker executes hidden side tasks inside deterministic Python sandboxes, while a Qwen2.5-1.5B LoRA monitor is trained with TRL GRPO and verifiable rewards to flag sabotage from public tool-call traces, reaching 0.893 AUROC with 0.88 recall and 0.12 FPR.

Built an OpenEnv reinforcement-learning environment where agents test a deliberately buggy REST task-management API. The environment includes 13 planted vulnerabilities mapped to OWASP API Security Top 10, seed-randomized data, deterministic bug detectors, a 5-signal reward function, and automatic OWASP-style bug bounty reports.

Implemented the complete DeepSeek V3 architecture from scratch, a 100M+ parameter transformer featuring Multi-Head Latent Attention (MLA), Mixture of Experts (MoE), and Multi-Token Prediction (MTP). Trained on the FineWeb-Edu dataset with ~2.5B tokens on an NVIDIA A100 80GB GPU.

Fine-tuned various open-source LLMs including LLaMA 2, Mistral, Qwen, and vision-language models for domain-specific tasks. Leveraged efficient methods like LoRA, QLoRA, and quantization using Unsloth and Hugging Face.

Implemented a Large Language Model (LLM) from scratch, covering every stage from data preparation and model architecture to pretraining and fine-tuning. This project demystifies transformer-based models through hands-on code and experiments, enabling a deeper understanding of attention mechanisms and token prediction.

Built an end-to-end MLOps pipeline to predict the approval status of US visa applications. Implemented machine learning models, deep learning techniques, and automated deployment pipelines.

Developed an end-to-end MLOps project to detect malicious URLs using XGBoost. Integrated robust pipelines for data ingestion, model training, deployment, and monitoring.

Predicted customer satisfaction scores for future orders using historical e-commerce data from the Brazilian E-Commerce Public Dataset by Olist. This project leverages multiple machine learning models like CatBoost, XGBoost, and LightGBM, built within a ZenML pipeline to create a production-ready solution.