AI/LLM Engineer · Forward Deployed Engineer
Almaty, Kazakhstan · sabit-shaikholla.github.io · LinkedIn · [email protected]
I build retrieval and agent systems over messy internal data: 100+ repositories, 1000+ Confluence pages and 5000+ documents, most of them scanned PDFs and badly formatted DOCX that break top-k retrieval. Evaluation runs inside the loop instead of after it (RAGAS, 0.89 faithfulness and 0.92 answer relevance on a 15K-query benchmark). I also serve models on-premise: Kimi-K2.6 on SGLang at 100+ tokens/sec, GPT-OSS 120B on vLLM.
Before the LLM work: Spring Boot microservices, event-driven backends, edge and IoT, plus a stretch leading ten engineers on platforms with 3M+ MAU. Most of it customer-facing, figuring out what someone needs before building it.
Repos are below. Write-ups are on my site.
Python, Java, Go · Google ADK, LangGraph, PyTorch, Vertex AI · vLLM, SGLang · OpenEvals, LangSmith, MLflow · Spring Boot, FastAPI, gRPC · Kafka, PostgreSQL, pgvector, Elasticsearch · Docker, Kubernetes, AWS, GCP
Open to AI/LLM and forward-deployed engineering roles. Email is fastest: [email protected].