Real jobs, done with tools you run yourself.
Pick a job below. Each one shows the problem, how the tool handles it, and the real screens or flow behind it.
Job 1 of 9with Nias
Answer data questions without shipping your data anywhere.
Without it
Building a chart usually means copying query results into an external analytics service, and then nobody is sure who saw what.
With Nias
Nias keeps exploration, SQL, saved queries, dashboards and exports in one self-hosted workspace, with permissions and an audit trail around all of it.
Read the Nias docsQuery workspace
Browse schemas, preview tables, write SQL, inspect results, save reusable queries and review history in one place.
Dashboards
Turn results into tables, KPIs and bar, line, area, scatter, pie or donut charts.
Exports and sharing
Export to PDF, PNG, Excel, CSV, SQL or JSON, publish read-only links, or embed a whole dashboard or one block.
Governance
Users, roles, permissions, 2FA, approval workflows, audit logs, monitoring, backups and a scheduler.
Nias stores its own app data in SQLite or PostgreSQL, with optional Redis for caching and rate limiting.
Not sure which job is yours?
Tell us what you're trying to do and we'll point you to the right tool, or try one on your own machine first.
Email usSee the install commands Job 2 of 9with Aras
Decide who can touch production, down to the statement.
Without it
As a team grows, shared database credentials turn into a liability. Anyone holding them can run a DROP TABLE on production.
With Aras
Aras layers access from the platform role down to a single user on a single instance, checks SQL against your rules as people type, and logs every query.
Read the Aras docs-
Role Which features you can use
Developers query and browse; DBAs also manage rules and changes
-
Access group Which instances your team sees
Team Alpha sees prod-db and dev-db
-
DB permissions Which statements are allowed
SELECT only on prod-db, full access on dev-db
-
Direct assignment A cap for one person
Alice: SELECT only on analytics-db
DROP TABLE orders;
Blocked by rule “No DROP on production”. Nothing ran.
SQL audit rules, checked live in the editor
- Block Block DROP TABLE, TRUNCATE, or DELETE without a WHERE clause on production
- Warn Warn on SELECT * against large tables and nudge people toward explicit columns
- Toggle Turn rules on or off instantly, with no deployment
Every query run through Aras is recorded: who ran it, on which instance, how long it took, how many rows, and whether it worked. Admins see everyone; developers see their own.
Job 3 of 9with Aras
Move risky SQL from chat approvals to a reviewed pipeline.
Without it
Schema migrations and bulk updates on production often get approved with a thumbs-up in chat, with no record of what actually ran.
With Aras
Aras takes every change from draft to dry run to a matched approval chain to execution, and writes each step to the audit log.
- 1
Draft
A developer writes the SQL, picks the instance and database, and adds a title and description. Everything stays editable.
- 2
Dry run
Aras runs the SQL inside a transaction that is always rolled back, returning errors, rows affected and timing without persisting anything.
- 3
Review
The approval chain matched to the instance and the developer's group assigns reviewers, for example Tech Lead, then DBA, then Security.
- 4
Execute
Once every step approves, an authorised user runs it. The result is stored with the full history and the actor's identity.
Reviewers only see changes for instances they can access, creators see their own, and admins see everything. Auditors get a complete, reviewable record of who approved what.
Job 4 of 9with Vestra
Stop opening a different console for every cloud.
Without it
Mixing GCS, S3-compatible stores like R2 and MinIO, and regional clouds like Huawei OBS or Alibaba OSS means hopping between consoles all day.
With Vestra
Vestra puts every bucket in one interface. Add the credentials once, they stay encrypted on your machine, and every file operation works the same everywhere.
Read the Vestra docs| Provider | What you paste |
|---|---|
| Google Cloud Storage | Service account JSON key (none for public buckets) |
| Amazon S3 | Access key, secret and region |
| Cloudflare R2 · MinIO · Wasabi · DigitalOcean Spaces · Backblaze B2 | S3-style keys plus the provider endpoint |
| Azure Blob Storage | Account name and account key |
| Alibaba OSS · Huawei OBS | AccessKey pair plus the regional endpoint |
| Google Drive | Service account JSON key and a Shared Drive folder ID |
- Same folders, breadcrumbs, previews and uploads for every provider
- Edit Content-Type, Cache-Control and custom metadata without re-uploading
- Test a connection against the live provider before you save it
with Vestra
Give MinIO and R2 the interface they deserve.
Without it
Self-hosted object stores are cheap and fast, but their consoles are heavy or awkward for everyday file work.
With Vestra
Add MinIO or R2 to Vestra as an S3 connection with a custom endpoint, and browsing, uploads, renames, bulk deletes and metadata edits just work.
MinIO credentials {
"access_key_id" : "minioadmin" ,
"secret_access_key" : "minioadmin" ,
"endpoint" : "http://localhost:9000" ,
"region" : "us-east-1"
}
- Run it next to MinIO as a single binary or a Docker container
- Connect several MinIO servers or R2 accounts side by side
- Object count and total size for each bucket, with instant search
with Mikko
Follow an attacker, not a wall of alerts.
Without it
The stock Wazuh dashboard shows events one by one. Spotting a campaign, opening a case and blocking an IP means stitching tools together.
With Mikko
Mikko pivots on the attacking IP, correlates the live stream into campaigns and floods, tracks the incident as a case, and blocks through Wazuh only when you arm it.
Read the Mikko docsThreat hunting
Ranked attackers, per-IP profiles with timelines, targeted agents, MITRE techniques, raw payloads and VirusTotal or AbuseIPDB links.
Correlation engine
A background task groups the live event stream into campaigns, floods and surges, and can auto-block once armed.
Case management
Cases with observables, append-only timelines and status workflows, opened straight from a detection or a hunt.
Guarded response
Blocks go through Wazuh firewall-drop. Dry run by default, allowlist-protected, every action audited.
Mikko never re-indexes Wazuh data; alerts, agents and vulnerabilities are fetched live.
Job 7 of 9with CLI
Use any AI model from your terminal, and keep a hand on the brake.
Without it
Every AI provider ships its own CLI, and agents that can edit files and run commands are hard to trust on a real repository.
With CLI
Anveesa CLI puts OpenAI-compatible APIs, local models and command providers behind one prompt, reads freely, and asks before it writes or runs anything.
Read the CLI docsOne command
Interactive chat, one-shot prompts and stdin review, for example piping a patch in for a second opinion.
Your choice of model
OpenAI, OpenRouter, Sumopod and other hosted gateways, local servers like Ollama or LM Studio, or Claude Code, Codex and Copilot CLI.
Read tools
List directories, find files, search text and read capped snippets. Secret files like .env and SSH keys are blocked.
Approval modes
writes:ask confirms each write or command, --yes allows them for one run, and stdin runs keep writes off by default.
Job 8 of 9with Workflow
Sketch the system with your team, without anyone signing up.
Without it
Whiteboard tools want accounts, seats and installs before the first box is on the canvas.
With Workflow
Open Workflow in a browser, share the room URL, and everyone draws on the same canvas live. Boards stay in each browser.
Read the Workflow docs- 1
Start a board and share the URL
Everyone who opens the link joins the same live canvas, with their own cursor.
- 2
Use the diagram palettes
Flowchart and Activity shapes with connectors that route and label themselves.
- 3
Iterate freely
100-step undo and redo, copy, paste and duplicate. Reopen the URL later and the board is still there.
- 4
Export a PNG
Drop it into a ticket, a wiki page or a pull request description.
Workflow is free, MIT licensed and needs no account. Nothing leaves your browser unless you export it.
Job 9 of 9with Aras, Vestra, Nias
One feature release, three tools, no SaaS in between.
Without it
A release touches the database, the static assets and the numbers people watch afterwards, usually across three vendors.
With Aras
Aras, Vestra and Nias each own one part of the release and all of them run on your infrastructure.
- Aras Migration approved
- Vestra Assets uploaded
- Nias Dashboard checked
- Nias Watching the numbers
- Aras
Prepare the database change
Create a change request with the ALTER TABLE migration, dry run it, and send it through approval. The DBA approves.
- Vestra
Upload the new assets
Upload the new JS bundles and images to the production S3 bucket or R2, updating Cache-Control in the same step.
- Nias
Validate data and dashboards
Check the migration results, update a saved query and refresh the release dashboard, then share it read-only if stakeholders need it.
- Nias
Watch and respond
Follow dashboards, query performance, audit trails and notifications. If numbers look wrong, the approved correction path is ready.