AI platform engineerBacked by humans

Dave gets infrastructure done for ambitious R&D teams.

2500+Dev tasks automated with
Dave (and counting)

Ask in Slack. Dave investigates and plans the change, a Dave engineer checks it, and your team approves before anything ships.

  • AICPA SOC 2
  • GDPR
  • HIPAA
  • ISO 27001
#kitefin-x-daveExternal
  • Compliance: The CTO asks what is still open before the SOC 2 observation window. Dave maps 7 failing Vanta tests to 4 Terraform PRs, a Dave engineer handles the CI user exception, the CTO approves, and Dave re-verifies all tests in Vanta.
  • FinOps: The platform lead asks why AWS jumped. Dave traces spend to unused NAT gateways, an oversized staging database, and detached volumes, a Dave engineer reviews the plan, the platform lead approves, and Dave verifies against live state.
  • Identity: A developer asks for a prod IAM key for Bedrock. Dave proposes a scoped, one-hour permission set instead, a Dave engineer confirms the scope, the platform lead approves, and the access goes live without a static key.
  • Upgrades & Migrations: The CTO asks what breaks in an EKS 1.29 to 1.30 upgrade. Dave finds one removed API in the ingress controller, drafts a staged wave plan, a Dave engineer reviews it, the CTO approves, and staging upgrades cleanly before the booked prod window.
  • Drift: The platform lead is about to apply a Terraform plan. Dave catches that it would recreate 14 existing resources, restores the missing import blocks, the platform lead merges and applies, and a Dave engineer verifies the post-apply plan is clean.
Manifold Labs
Manifold Labs
Manifold Labs
Manifold Labs

Case study

SETT

How SETT cut infrastructure costs 20% and served agents twice as fast with Dave

SETT builds agentic AI for mobile game marketing teams. Its engineers now send new infrastructure work and cost questions to Dave, and spend their time on the product.

Read the case study
Cut in infrastructure costs after Dave traced a sudden spike
20%
Cut in infrastructure costs after Dave traced a sudden spike
Faster agent serving after a platform change planned with Dave
~2×
Faster agent serving after a platform change planned with Dave
Of engineering time each week that infrastructure work used to take
~12 hrs
Of engineering time each week that infrastructure work used to take

Context first

Dave reads before it writes

One Slack message, every place the answer could be hiding. Dave checks your live infrastructure, recent changes, Terraform state, past incidents, and runbooks before it proposes anything.

Every answer traces back to what Dave actually looked at, and the fix still waits for your approval.

Chris Delgado1:02 PM
@Dave staging deploys for payments started failing after lunch. Can you look?

Dave checked

  • KubernetesLive cluster state

    payments-api CrashLoopBackOff

  • GitHubRecent changes

    PR #588 merged 12:41 · bumped base image

  • DatadogMetrics

    5xx rate up from 12:44

  • Terraform state

    No drift · infra unchanged since Monday

  • SlackPast incidents

    Mar 14 · same error, fixed by pinning libssl

  • Runbooks

    payments-api: rollback and image pinning

Dave APP1:09 PM
PR #588's new base image dropped libssl 1.1, the same failure as the March 14 incident. I drafted PR #591 to pin it, plus a rollback if you'd rather revert.Proposed fix · needs your approvalReviewing: Danielle (Dave human engineer)

Plan first

Look before you leap

Every Dave change starts as a plan, not a commit. You see what changes, what it could affect, what it costs, and how to undo it before anything runs. The plan is the contract.

Plan · kitefin / infra · PR #318

Remove unused NAT gateways and resize staging-pg

Waiting for approval

Goal

Cut spend left over from the May worker migration without touching production traffic.

What changes

  • aws_nat_gateway ×2 in the old worker VPC · delete
  • staging-pg · db.t4g.medium → db.t4g.small, Multi-AZ off
  • aws_ebs_volume ×11 unattached · snapshot, then delete

Blast radius

Staging only. No production resources. The NAT gateways moved 0 bytes in the last 30 days.

Cost

Proposed −$299/mo · verified after it lands

Security & compliance

No IAM changes. Snapshots kept 30 days to match the SOC 2 retention policy.

Rollback

Restore volumes from snapshots; recreate the NAT gateways from Terraform. Resize is reversible in the next window.

Review & approval

Reviewed by Danielle (Dave human engineer)

Anatomy of a Dave plan

Try it: approve the plan and watch the proposed cost become a verified one.

Zero platforms. Zero workflow building. Just outcomes.

  • Nothing new to adopt

    Dave works in Slack and in the repos, clouds, and Terraform you already run. There is no new platform, portal, or dashboard for your team to learn.

  • No workflows to build

    Ask in plain language. Dave works out the steps, the tools, and the owners, so nobody writes a runbook or wires up an automation first.

  • The finished change, not advice

    You get a reviewed PR, an applied change, and a record of what was verified, instead of a list of suggestions to carry out yourself.

Backed by humans

Every change has a human engineer behind it

Dave does the legwork. A Dave engineer in your time zone joins your shared Slack channel and stays accountable for the calls that need judgment. Your team still approves every change.

In the demo above, Danielle is the
Dave human engineer on Kitefin's channel.

  1. 01

    Reviews every plan

    Before your team sees a plan, a Dave engineer checks the scope, the risk, and the rollback path.

    Danielle Brooks(Dave human engineer)10:47 AM

    Reviewed. Snapshots before any delete, resize in tonight's window. Good to approve.

  2. 02

    Takes the judgment calls

    Shared credentials, unclear ownership, production risk: when a change needs judgment, Dave hands it to an engineer instead of guessing.

    Danielle Brooks(Dave human engineer)9:14 AM

    Taking the CI user: moving deploys to GitHub OIDC, so there's no key at all.

  3. 03

    Signs off on the result

    After the change lands, an engineer confirms the live state matches the plan and closes the loop.

    Danielle Brooks(Dave human engineer)4:44 PM

    Verified state and locking on the networking stack. Closing this one out.

Who Dave is for

For engineering leaders whose infrastructure work outgrew the team

No infra teamFor CTOs and VPs of Engineering

Your developers are doing the infrastructure work

Cloud tasks land on whoever knows AWS best, and product work waits. Dave takes the infrastructure queue, a Dave engineer covers the judgment calls, and your developers get back to shipping.

  • Access, environments, and cost questions handled in Slack
  • SOC 2, upgrades, and migrations carried to done
  • Nothing ships without your approval
Lean infra teamFor Heads of Platform and DevOps leads

Your platform team can't keep up with the queue

When one infrastructure engineer supports ten or more developers, the backlog always wins. Dave clears the requests and the recurring work, so your team spends its time on the platform.

  • Works in your Terraform, pipelines, and approval path
  • Recurring checks that actually run
  • Your team keeps production authority

Hand Dave your next infrastructure task

Bring the thing that's been stuck in the backlog. We'll show you how Dave and a Dave engineer would get it done in your stack, with your team approving every change.

Trusted by R&D teams.

“Our team clearly needed the backup, and hiring someone new was proving to be a nightmare. dave.io a shot eight months ago and we're still with them today. The work gets done, it gets done right, and I'm not babysitting anyone. I see us continuing with them for the foreseeable future”

Tomer Kulla

Chief Technical Officer

“Dave has been excellent. The move from ClickOps to GitOps was smooth, thanks to their hands-on approach and DevOps expertise. They provided recommendations, supported implementation and took ownership of what was needed to move things forward. Their proactive approach helped us build scalable, maintainable infrastructure. It feels less like a vendor and more like an extension of our team.”

Dori Lahav Waisberg

Full Stack & DevOps Engineer

Manifold Labs

“I work in AI so I tend to be skeptical of most 'AI-powered' anything. dave.io is the exception. It's not magic, it's just competent. Infra stays safe, deploys don't break, and when I have a weird request it gets figured out without a back-and-forth because at the end of the day, I'm dealing with a real person, not just an agent. Good product. Recommended.”

Yinon Goldstein

AI Researcher

“Does what it says. Team's responsive, work is clean, no hand-holding required. We're a startup, everyone's stretched. Last thing I need is another vendor I have to manage so I appreciate that dave.io just runs. Issues get caught, things get built, and I don't hear about problems unless I need to. Honestly just feels like having a reliable person on the team. That's rare, and I don't say it lightly.”

Yuval Hasson

CTO & Co-Founder

Xpend Marketing

“We scaled fast and our team was struggling with capacity. Dave came recommended, so we went for it. I've worked with many vendors, and most need managing. Dave feels less like a vendor and more like a remote team member. The work is consistent, communication is straight, and collaboration with my team makes sense. I’d recommend it to any fast-moving R&D team looking for an alternative to full-time hires.”

Rotem Bloom

VP R&D

Integrates with all your tools.

AWS
Kubernetes
Terraform
GitHub
GitLab
Datadog
Vanta
Okta
Slack
Linear
Docker

Safe, reliable, and always on

keeping infrastructure work secure and dependable around the clock

> 99.9% Uptime24/7 coverage
AICPA
GDPR
HIPAA
ISO

Learn what Dave would find in your stack

Skip the research and let ChatGPT, Claude, or Perplexity do it for you. One click shows you what Dave can do.

Dave's ready to build.Are you?

By submitting this form you agree to the processing of your personal data as described in our Terms and our Privacy policy.