Projects
Every project below was built solo.
WarpQuantA dual-domain post-training quantization method that compresses transformer weights after Hadamard rotation and spends a small recovery budget on columns ranked by Output-Fisher sensitivity. The same report evaluates a TurboQuant-style KV cache and per-token INT8 activations.
Uncertainty-Aware Bid MLA q1000 and Monte Carlo decision system that provides bid-win probabilities and uncertainty for each client.
Document AI OCRA deterministic Document AI pipeline that structures HWP, HWPX, and PDF notices while separating field copying from amount, ratio, and unit validation.
Procurement NLPA CPU NLP API that classifies bid eligibility and detailed work categories from Korean procurement-notice titles.
R2CCP Bid PredictionA multimodal bid-rate forecasting system that preserves disjoint conformal regions and turns eight context distributions into candidate decisions through 500K Monte Carlo simulation.
google-surf-mcpAn MCP search server that automatically stores web, paper, and code search results in a local knowledge graph, then retrieves them alongside fresh sources through hybrid RAG.
MonogramA personal knowledge system that turns one share action into organized, searchable knowledge from news, social media, papers, videos, and documents.
Bau BrowserA local-first desktop browser where users control private data and agent actions through a scoped propose → approve → execute flow.
The Pre-Execution Identifiability Bottleneck in Fixed-Pool LLM Agent Configuration SelectionResearch asking whether task text, embeddings, or hidden states can identify the best LLM-agent configuration before execution.
EMH AgentA reinforcement-learning investment agent that builds and evaluates portfolios using only the information available at each point in time.
MyShotConnects golfer, club, and ball tracking with 3D swing mechanics and personalized shot-outcome prediction from one phone video.
EATE-AT combines focal loss, bidirectional R-Drop, and adaptive FGM so an ISIC skin-lesion classifier learns both class quality and better-calibrated confidence.
WSSSImproves segmentation learned from image-level labels by finding unreliable pixels in automatically generated masks and repairing only those regions before retraining.
LangGraph Travel AgentA travel advisor delivered to a U.S. agency that turns a natural-language request into three bookable packages and connects review, CRM, and customer messaging.