Where the business meets engineering.
Intake, routing and agent orchestration for teams building with AI.
curl -fsSL sleds.ai/install | shSSO redirect loop on the Acme tenant
For the business
Request a feature, an update or a fix from the AI you already use. Every request arrives understood: the why, the evidence, the deadline, the worth.
Intake →The core
Sized from your team's history, then sent to the right engineer, agent and model.
Routing →For engineering
Engineers and their agents deliver it, on their machine or in the cloud, and every result is signed and measured.
Orchestration →Your AI, now fluent in your codebase.
Request a feature, an update or a fix in Claude, ChatGPT or Cowork, whichever you already use. It knows what the request touches, asks the two questions the spec missed, mocks the change and files it ready to build.
Sized from your history. Sent where it fits.
Every request gets a size before anyone starts, drawn from the work your team has already landed in that part of the code and from how your CI behaves there: build times, first-try pass rates, the flaky suites. Each receipt makes the next size better. Then sleds picks the route.
Your rules decide what goes out on its own. Everything else arrives as a recommendation, with the reasons, for someone to take.
One afternoon on the sled.
Follow one request, ENG-482, from Samantha's chat to John's Friday digest. The steps play on their own; pick any to stop there.
Samantha requests a fix from her Claude.
No ticket, no login. She says "send this to engineering" with the spec she just wrote. Her Claude, running our instructions, asks the two things the spec did not answer (what does done look like, and what is the smallest version worth doing) and plays the brief back. She says yes.
Intake, in full →A band, not a guess.
Before anyone commits, the request gets a letter, a dollar and minute band, a confidence and a cap, from 112 rows this team landed in the same area. Which models fit is on it too.
How sizing works →size M $4–11 20–55 min 81% basis 112 rows · auth · this repo fits Fable (band) · Sonnet (+40% time, ⅕ cost) · Haiku does not fit worth high · Acme renewal (stated by Samantha) cap $20 · nothing bills past it suggest now · Jake · 62% of the weekly bar · resets Thu 9:00 on the row ENG-482 · Samantha told in her thread
Staged to go, with a suggestion.
Every request, sized and valued, waits on the board with a suggestion: now and to whom, overnight at half price, or hold and ask the filer. A lead reads it in a glance.
The board →| request | size | worth | suggest |
|---|---|---|---|
| SSO redirect loop · Acme | M | high | now · Jake · headroom |
| Rate-limit headers · API | M | — | hold · ask the filer |
| Export CSV filter | S | medium | overnight · ½ price |
| Empty state · new workspace | S | low | overnight · ½ price |
The wall is an exit, priced six ways.
Jake hits his weekly limit mid-session. The session is snapshotted and he sees six exits with prices on them, in his own terminal. Policy says hold, unless worth is high.
Routing →Wall: usage limit reached. Session snapshotted. Provider: Weekly limit reached. Resets Thursday 9:00 am. Remaining work: $4.10–9.30 · 20–55 min · 82% 1 Continue now platform key · allowance left $38.00 2 Hold until reset free · resumes 9:00 am EDT · saves ≈ $6.70 vs continuing now 3 Run overnight batch tier · ½ price 4 Switch model sonnet 5 · platform key · $0.90–2.10 · same time · rest of session 5 Hand off to Dana (idle 42m) · Slack + /sled pickup 6 Defer to reset runs itself 9:00 am EDT · same model · full price · $4.10–9.30 > /sled 3 Running overnight. Capped $20 · batch tier · ½ price. Delivered by 6:20 am.
A receipt on the pull request.
On merge, the PR carries its source, worth, estimate, seat, actual and a signed receipt. The row moves to Done, Samantha gets her receipt, and the next size is a little better.
The record →SLEDS-Source: ENG-482 · brief · the board Worth: high · Acme renewal SLEDS-Estimate: $4–11 · 20–55 min · 81% · cap $20 SLEDS-Seat: overnight · batch · fable-5.1 SLEDS-Actual: 31 min · $5.20 SLEDS-Attest: ed25519 · 2026-09-14T06:41Z · verified → ENG-482 moved to Done · receipt sent to Samantha · the next size is a little better
The week, in the channel.
John reads the roll-up where the team already talks. What landed, what is waiting and on whom as a queue, what hit a wall. He adjusts worth on two rows.
Engineering leads →This week on the sled illustrative landed 47 rows · $212k worth · $1,140 cost in band 78% of sizes · calibration moving up overnight 19 rows at half price · $71 saved walls 12 hit · 12 priced · 0 waited review billing is the queue · 2 days median unvalued 5 rows · ask the filer
You don't need a tracker anymore. Or a wiki.
The conversation is the form. The board is the record: every request with its brief, size, worth, seat, questions, prototype and receipt. Docs holds the specs and decisions beside the work, and a doc knows when the item it describes has landed. GitHub carries the code. Slack carries the news. Moving off an old tool? Pick a day and it becomes read-only history.
Nobody learns a ticket template again.
Priced before, signed after, one row each.
Specs and decisions, tied to items. Import a Confluence space in one run.
The code, the receipt, the news.
Everyone who touches the work, in one place.
Business teams
Request what the customer needs from the chat you already have open. No seat, no login. Free for everyone, forever. Hear when it lands, with what it cost.
Business teams → product · designProduct teams
Specs and prototypes that arrive ready to build. Worth set by the people who know what it's for, and a view of what's in flight, waiting and landed.
Product teams → engineers · leads · the ctoEngineering
The right work on the right seat and model. A wall that is an exit. Value in flight, landed and waiting, by area.
Engineering →We're bringing teams on a few at a time.
Leave your work email and we'll get in touch to set your team up. Start with one engineer: useful the same day, and it costs the workspace nothing.
curl -fsSL sleds.ai/install | sh