Fine-Tuning LLMs at Scale with Databricks MLflow and Spark
11 min read4.7k
> From raw data to a registered, served fine-tuned model — a production walkthrough using Databricks, Apache Spark, MLflow, and Hugging Face Transformers. --- T...
#LLM
7 posts
Benchmarking Foundry IQ Against a Custom Retrieval Pipeline in Microsoft Foundry
6 min read3.4k
Foundry IQ is Microsoft Foundry's managed knowledge-base layer, built on Azure AI Search under the hood, adding automatic freshness handling and simplified hybr...
Designing Reliable Function-Calling Schemas for Production in Microsoft Foundry
6 min read4.4k
Function-calling demos work because the model is well-behaved and the test queries are clean. Production breaks this in three specific ways: malformed arguments...
AI-First Backend (RAG + APIs + Caching)
7 min read5.1k
In the traditional world of distributed systems, our primary concern was the deterministic flow of data: a request comes in, we query a relational database, apply business logic, and return a JSON res...
AWS Bedrock vs Self-Hosted LLMs: When to Choose What
5 min read7.1k
The shift toward Generative AI has forced cloud architects to move beyond traditional CRUD applications and grapple with a fundamental "Buy vs. Build" dilemma: should we leverage a managed service lik...
LLM Inference at Scale (ChatGPT-Style Architecture)
7 min read6.1k
Building a production-grade system for Large Language Model (LLM) inference at scale represents a fundamental shift in distributed systems design. Unlike traditional microservices at companies like Ub...
Hosting LLMs on AWS: ECS vs EKS vs SageMaker
6 min read5.9k
The rapid proliferation of Large Language Models (LLMs) like Llama 3, Mistral, and Falcon has shifted the cloud engineering focus from model training to efficient, scalable inference. For organization...