BucketDesk
Sign inStart free trial
Engineering8 min readAugust 31, 2026

A cost conscious technical stack for an S3 file workspace

Start with a modular application, browser native previews and one isolated worker. Add managed media and search infrastructure only after demand justifies them.

JeVaughn Ferguson
Founder, developer
The short version

Ship secure S3 access and browser native formats first. Make expensive processing asynchronous and replaceable, then scale the component that becomes a measured constraint.

The cheapest credible architecture is not a collection of free libraries. It is a small system with strong boundaries and a clear path to replace components as usage grows.

For an initial release, one control plane and one isolated processing worker are enough.

Keep the application together

  • Laravel provides authentication, policies, queues, notifications and administration.
  • Inertia, React and TypeScript provide a rich file interface without maintaining a second API application.
  • PostgreSQL stores organizations, permissions, object metadata, workflows and audit history.
  • The Laravel database queue is sufficient until job volume requires SQS.

Use the browser before building a conversion service

  • PDF.js for PDF files.
  • Native browser rendering for supported images, audio and video.
  • Sanitized Markdown, text, JSON and CSV viewers.
  • hls.js when adaptive HLS playback is required outside native support.
  • Short lived delivery access with byte range support for media seeking.

Isolate the tools that process customer files

A separate container can run LibreOffice, FFmpeg, archive tooling, thumbnail generation and malware scanning. Each job needs temporary storage, resource limits, a timeout, restricted network access and a durable result.

Let customers fund provider work where practical

AWS STS can provide temporary credentials for a customer role. The role and any session policies determine what the session can access. Original files and most generated output can remain in customer controlled storage.

Defer infrastructure with no evidence yet

  • Do not begin with Kubernetes, OpenSearch or a distributed workflow engine.
  • Do not transcode every media file automatically.
  • Do not index every document for AI before the customer enables it.
  • Do not support every archive or object store in the first release.
  • Do not add a commercial viewer until its fidelity saves more engineering cost than its license consumes.
Try it in BucketDesk

Minimum is free. Deploy a scoped role with CloudFormation, sign in, and browse, without handing anyone an access key.

Connect a bucket

Primary sources

Read next
Article 2 of 97 min

AWS FinOps Agent finds the cost. Knowing what to fix, and fixing it, is the hard part

The FinOps Agent tells you why the bill moved and lists what AWS recommends. Someone still has to choose the right change, make it without breaking anything, and check next month that the money actually came off.FinOps Agent preview scope (Cost Anomaly Detection, CloudTrail correlation, Cost Optimization Hub and Compute Optimizer summaries, Jira and Slack delivery), the hand-off point, and how BucketDesk's S3 Savings tool covers assessment, one-click correction and verified outcomes with timelines.

Discussion

0 comments · open to guests · moderated
Comments appear after a quick review.

Liked this? Get the next article by email. No schedule, no filler, one click to leave.

Keep reading

All writing →
Product decisions9 min

Amazon Quick Suite for documents in S3: what it indexes, what it costs, and when you don't need it

AWS now sells Amazon Quick Suite instead of Amazon Q Business. If you want it to answer questions about files in your bucket, here is what you sign up for and what it costs.Quick Suite tiers and the $250 infrastructure fee, pooled index GB and $5/GB overage, the Enterprise prerequisite for S3 knowledge bases, same-Region and ACL-at-creation constraints, 500 MB file and 30 MB extracted-text limits, daily sync, and the per-document alternative.

Operations9 min

Amazon S3 storage classes: which one for which business file

Every file in S3 sits in a storage class that decides what it costs and how fast it opens. Here is which class fits which kind of file, in plain terms.Class codes, the two deciding questions, a per-file-type map, --storage-class on upload, copy-in-place to change class, a waterfall lifecycle rule with ObjectSizeGreaterThan, the 128 KB default, the 94-day ordering rule, and why reads keep Intelligent-Tiering objects warm.

How to8 min

How to reduce Amazon S3 costs: the three settings that cut most bills

The S3 bill is usually high for three boring reasons. Each one is a setting on the bucket, and each one can be checked and fixed without touching a file.NoncurrentVersionExpiration with NewerNoncurrentVersions, AbortIncompleteMultipartUpload, Intelligent-Tiering with optional archive tiers, the 128 KB rule, worked numbers at us-east-1 list prices, and how to read Cost Explorer by usage type.

We use cookies and similar technology for a few things, like:If you choose “Accept all”, we’ll also use Google Analytics and Google Ads cookies for things like:

We never send file contents, credentials, email addresses, or secure link tokens.

Read the privacy notice