500 free credits, no card. Try Smalt AI →

Showcase

See the work.
Not the promise.

Anyone can generate a slide. The question is whether the output survives the person who has to use it. Here is what your AI coworker hands back from a one-line brief: working files, sourced claims, and the steps it took.

The platform in action

Every capability, in motion.

One plain-language brief goes in. A spreadsheet with live formulas, an editable deck, and sourced research come out — watch it happen.

Smalt AI one brief, finished work
Plain language in

northwind-brief.docx

Word · 2 pages

“Prep me for tomorrow’s renewal call with Northwind.”

What came back

  • The account and open deals from HubSpot, plus who is attending and their roles
  • Recent company news, every claim linked to its source
  • Follow-up email written and waiting for your approval

Shows its steps: Read the account → looked up 4 attendees → checked the news → drafted the follow-up

vendor-reconciliation.xlsx

Excel · reconciliation

“Reconcile last month’s vendor invoices against the PO sheet.”

What came back

  • Matches built with =XLOOKUP — change a PO and the match updates
  • Every exception listed and linked to its invoice
  • Totals by vendor that recompute on edit

Shows its steps: Read the invoices → matched against POs → listed the exceptions

q3-campaign-review.pptx

PowerPoint · 12 slides

“Turn Q3’s campaign numbers into a review for Friday.”

What came back

  • Native editable .pptx — real charts, not screenshots
  • Spend, signups and cost per signup by channel
  • On your template, watermark-free

Shows its steps: Read the campaign sheet → found what moved → laid out 12 slides

competitor-launches.pdf

Research brief · sourced

“What did our three main competitors launch this quarter?”

What came back

  • Every claim links to the page it came from
  • Pricing and positioning changes side by side
  • Anything it could not confirm is marked open, not guessed

Shows its steps: Swept their sites, press and reviews → compared the launches → cited every claim

onboarding-tracker.xlsx

Tracker + 3 email drafts

“Get Monday’s three starters sorted. Checklist is in Notion.”

What came back

  • Three welcome emails and a Slack note to each manager, waiting for approval
  • Tracker updated with what is done and what is waiting
  • One start date that did not match the offer letter, flagged

Shows its steps: Read the checklist and hiring sheet → drafted the emails → updated the tracker

refund-procedure.docx

Process doc · Notion draft

“Write up how we handle refunds, from the refund emails and the checklist.”

What came back

  • Step-by-step procedure with an owner for each step
  • Where the emails and the checklist disagree, called out for a decision
  • Saved to Notion as a draft for review

Shows its steps: Read the emails → merged them with the checklist → flagged the conflicts

These are representative of the deliverables Smalt AI produces. Want to see one on your data? It takes a one-line brief and a few minutes.

Under the hood

The strengths of many models, for each step.

One chat. Behind it, each step goes to whichever model suits it, open-source or frontier. You never pick a model, and you never see the seams.

Reasoning & structure
Reports, decks, reconciliations — the work that has to be right.
Long-document analysis
Contracts, handbooks, long threads, hundreds of pages — read in one pass.
Fast classification
Sorting an inbox, tagging files, routing tickets — speed where speed is fine.

Not locked to one vendor. When a better model arrives, we adopt it, and your workflow does not change.

See it on your work.

Prep a call, reconcile a spreadsheet, build a deck. 500 free credits, no card.

Try Smalt AI See more use cases