---
title: "BentoML | Skills We Assign For | Azendo"
description: "BentoML for model serving — packaging, adaptive batching, plus the Azendo roles assigned for it."
url: "https://azendo.co/skills/bentoml/"
---

[Skills](https://azendo.co/skills/) MLOps and model delivery 

# BentoML.

BentoML packages trained models into standardised, containerised services with an HTTP API, handling dependency capture, adaptive batching and deployment to several targets.

## Where BentoML fits on a long engagement.

Packaging is the step where models most often stall. Capturing the exact dependencies, the preprocessing code and the model artefact into something reproducible is fiddly, and doing it by hand per model is how "it worked in the notebook" becomes a deployment blocker.

Adaptive batching arrives built in, which matters because implementing it well is non-trivial and it is one of the largest throughput levers available on GPU serving.

## What an assigned team does with BentoML.

The preprocessing and postprocessing around a model are where serving bugs live. Code that differs even slightly from the training path produces a model behaving differently in production for reasons that look inexplicable.

Packaging the transformation with the model rather than reimplementing it in a service is what prevents that, and it is the standard held under [managed ai services](https://azendo.co/services/ai-engineering/).

## What we use BentoML for.

* Reproducible model packaging Dependencies and preprocessing captured, removing the usual deployment blocker.
* Batching without implementing it Adaptive batching built in, which is a substantial throughput lever.
* One transformation path Preprocessing shipped with the model rather than reimplemented in a service.

## How BentoML capacity is assigned.

Model packaging is assigned inside AI capacity, with preprocessing shipped alongside the model to prevent serving skew.

## Roles we assign BentoML for

* [Model Deployment Engineer MLOps engineering](https://azendo.co/services/mlops-engineering/model-deployment-engineer/)

## Service lines it sits in

* [MLOps engineering](https://azendo.co/services/mlops-engineering/)

Capacity is agreed as a committed monthly capacity across a discipline, not per skill.

## Related in mlops and model delivery

* [MLOps handoff — skill we assign for](https://azendo.co/skills/mlops-handoff/)
* [ONNX — skill we assign for](https://azendo.co/skills/onnx/)
* [TensorRT — skill we assign for](https://azendo.co/skills/tensorrt/)
* [Model registry — skill we assign for](https://azendo.co/skills/model-registry/)
* [Weights & Biases — skill we assign for](https://azendo.co/skills/weights-and-biases/)
* [Comet — skill we assign for](https://azendo.co/skills/comet/)
* [Neptune.ai — skill we assign for](https://azendo.co/skills/neptune-ai/)
* [DVC — skill we assign for](https://azendo.co/skills/dvc/)

## Tell us what your roadmap needs BentoML for.

A service delivery manager replies with the disciplines we would assign, the monthly capacity and what the first month looks like.

[All skills we assign for](https://azendo.co/skills/)