AI Engineer | LLM & Applied AI
Repositories
BenvinD/lemma-rag
BenvinD/vortex-ai-gateway
Every team that calls an LLM ends up writing the same five things around the call: which vendor gets which model, a retry loop that does not make an outage worse, a limit on what one caller can spend, a cache for the question that was already answered, and a way to see what any of it cost.
BenvinD/claimlens
Multi-modal damage-claim evidence review. A claim arrives as a chat transcript plus photographs. ClaimLens decides whether those photographs support the claim, contradict it, or leave not enough information to judge — and returns a structured, fully enumerated row for every claim.
BenvinD/vllm
A high-throughput and memory-efficient inference and serving engine for LLMs
BenvinD/AdventCode2025
BenvinD/Crowd-Detection-And-Analyzation
A Python Program using OpenCV to detect the maximum numbers of people present in the frame and show visualised analyse of the data
BenvinD/MacOS_LLM_Server
BenvinD/Git_Workshop_2021
Make changes in this Repository and practice git.