Independent ML Researcher
Independent ML Researcher focused on NLP, LLMs, AI safety. I'm drawn to the unsolved parts of the field, where the science is still being written and the engineering hasn't caught up yet, and building the systems that make those ideas real.
Papers
-
CVPR, NTIRE
Education
Thesis:
Efficient Deep Single-Image Super-Resolution on Mobile
Devices
Supervisor:
Prof. Dr. Radu Timofte
A study on deep learning-based methods for efficient mobile
SISR across multiple upscaling factors (×2, ×3, ×4), optimizing
first for PSNR then perceptual loss. Architectures were kept
deliberately shallow to minimize inference time, compensating
through increased width to maximize GPU parallelism.
Thesis:
Resource Streaming using a Peer-to-Peer Architecture
Supervisor:
Prof. Dr. Lenuta Alboaie
Proposed a decentralized P2P resource streaming framework with
no imposed hierarchy, eliminating single points of failure and
reducing privacy risk. The architecture uses locality-aware
distributed hash tables (LDHTs) for per-node network state and a
custom epidemic protocol for network-wide information
propagation.
Experience
- Authored DEX (Data-Emergent Experts), a dense MoE-inspired architecture with an embedding-based router that clusters the training corpus by document meaning and assigns each cluster its own expert, weighted by softmax similarity to cluster centroids, enabling on-demand capability removal by zeroing target experts while preserving retained knowledge
- Authored ReViB (Recovered Visual Bias), a training-free contrastive decoding method that inverts an LVLM's visual encoding back into image space via unCLIP to recover and subtract the model's statistical bias, reducing object and attribute hallucination
- Authored Halfway Speculative Decoding, proposing joint drafter-target training via LK loss optimization with cross-entropy regularization on the target, achieving competitive draft acceptance metrics using only 9\% of the training data required by prior approaches
- Built agentic AI pipelines end-to-end, including monitoring and observability
- Played a major role in the company's agentic AI transition, contributing to technical direction and architecture decisions
- Implemented embedding model-based semantic indexing and retrieval systems
- Actively consulted on AI and data science decisions, shaping how the team approached and integrated intelligent systems
- Handled massive data volumes across distributed systems, ensuring data quality, feature reliability, and pipeline integrity that models directly depended on
- Actively consulted the ML team on causality analysis algorithms and contributing technical input throughout the model design and development process
- Managed complex DevOps workflows and large-scale deployment pipelines for global production systems
- Architected a scalable, production-grade system from scratch, owning the full pipeline from design through deployment
Skills
Certifications
