Data Engineer.
Learning AI
from Scratch.
I wrestle Spark clusters by day and spend my nights trying to figure out why PyTorch is angry at my GPU dimensions. I'm building AI systems from the ground up because calling a cloud API felt too easy.
Data Engineering
Building stream pipelines, optimizing database configurations, and fixing data quality checks at odd hours. This is my day job. Expect real production lessons here instead of neat tutorial code.
AI from Scratch
Writing tokenizers, matrix multiplications, and attention loops on paper before coding them. I want to understand why these models break under the hood, not just how to prompt them.
Real Applications
Moving beyond simple hello-world prompt templates. I build tools that process actual data pipelines and handle real constraints because the world does not need another basic wrapper.
NeuroLambda: How I Ran Two AI Models on One 8GB GPU to Build a Real-Time SRE Pipeline
Two models. One 8GB GPU. 707 minutes of continuous operation. NeuroLambda combines Mamba S6 (0.41ms/event) and Qwen-3B for automated root-cause diagnosis, achieving F1=0.9713 on zero-shot microservices data with 51% lower latency than static routing.
Why I Replaced Transformers with Mamba S6 for Log Analysis (70× Faster, Flat Memory at Any Length)
GPT-2 crashes OOM past 8,192 tokens and processes only 35 logs/sec. Mamba S6 runs at 2,453 logs/sec with constant memory regardless of sequence length. We built two 125M SSMs, fixed a 3GB streaming KV cache bug, and benchmarked both against GPT-2 across three random seeds.
I Trained a 124M-Parameter LLM to Catch Silent System Failures: Zero Labeled Data
No labeled failures. No regex rules. A 124M-parameter GPT-2 variant trained only on normal system logs detects anomalies at F1=0.8923 on a consumer RTX 3060 Ti by measuring how surprised it is by each log sequence.