AI agents built to learn how your sales and finance managers work, from how they structure a tough deal to which lender buyer they call, and put that know-how on every deal, every shift, inside the limits you set. They're built to know every program your lenders publish and work every approval into the most profitable deal that fits, before a manager is asked. Built to work on top of Dealertrack, RouteOne, CUDL and other credit platforms.
| Lender | Decision | Stips | Back-end room |
|---|---|---|---|
| Lender A | Approved | 2 | Low |
| Lender B | Countered | 1 | High |
| Lender C | Conditioned | 4 | Medium |
Three lenders answer overnight. Osprey reads each one against that lender's program, finds the fix and has the best deal waiting when Devon walks in.
Each agent takes one piece of the grind off your sales and F&I managers and works inside the parameters you set: markup, term, advance, products, which lenders. Anything outside them goes straight to your managers.
Penciling Deal 1042
Does the desking math so your sales manager can work more deals.
Never quotes a customer directly. Your desk presents.
Reading Bulletin 26-10
Knows every program from every lender you work with.
Never changes the customer's deal or crosses a limit on its own. That goes to a manager.
Sizing the menu to Lender B
Preps the menu so your F&I manager walks in ready.
Never packs a payment. Never adds a product without the customer's consent.
Checking proof of income
Takes stip chasing and contract checks off your F&I office.
Never alters a document or submits on its own.
Checking markup policy
Watches every step of every deal.
Not legal advice. Your compliance team owns the rules.
Robin is answering a 9:40 PM lead
Bring the deal in the door.
Never discusses price, payment or credit.
Agent activity shown is illustrative.
Lenders write their programs in bulletins, rate sheets and guides, each with its own definitions and fine print. Our agents are built to turn each one into exact, cited rules for that lender. Then they do the work: structure the deal for every lender that said yes, fix what doesn't fit, and rank the results by what the store keeps. Your manager gets the best deal, already checked, not a stack of approvals to rework.
| Rule | Deal | Result |
|---|---|---|
| max_term §4 | 72 mo · 2022 | Pass |
| max_advance §3 | 108.8% incl. back-end | Pass |
| pti_max p.3 §1 | $509/mo · 15.9% of gross | Over |
| backend_cap p.3 §2 | $2,950 vs $3,500 limit | Pass |
| special 26-14 | 11.9% (standard 12.4%) | Applied |
Sample deal: 2022 Camry SE, wholesale book $24,000, customer income $3,200 a month, 11.9% after Special 26-14. Every rule is cited to Bulletin 26-10. Runs in your browser; nothing is sent.
| Lender | Fix to fit the program | Back-end | Reserve | Stips | Store keeps |
|---|---|---|---|---|---|
| Lender BBest deal | +$1,500 down clears PTI | $2,950 | $610 | 1 | $3,560 |
| Lender C | None; conditioned approval | $2,400 | $540 | 4 | $2,940 |
| Lender A | None; as submitted | $1,800 | $480 | 2 | $2,280 |
Ranked by what the store keeps, then fewest stips, then lowest payment. All three stay inside the store's limits. Devon gets Lender B with the math behind it, and presents it because the down payment changes. Fictional figures.
Same store, same three approvals on Deal 1042. The difference is who does the math, and when.
One deal worked deeply. The rest get the first approval that fits.
Every deal worked the same way, before the morning starts.
Open one screen and see every deal waiting on a lender, every stip still missing, and every contract that hasn't funded, with the best deal already structured and the next move drafted.
Try it. These buttons only change this page. Nothing is sent.
Only Lender C approved, and only at 84 months, past your 75-month limit on used. Osprey laid out the deal against Lender C's program so you can decide.
Lender is holding funding for proof of residence. Request to the customer drafted.
Filled dots are the moments a person decides. Everything between them, the agents work inside your limits.
Robin answers a late-night inquiry. Finch proposes times for your team to confirm.
Kestrel suggests structures. The sales manager presents.
Osprey structures every approval to fit its lender and ranks them by what the store keeps.
Osprey handles counters that leave the customer's deal unchanged and stay inside your limits. Anything else goes to a manager.
Merlin builds it. Your F&I manager presents it.
Heron clears the stips and audits the contract.
We're early, so there are no results to show yet. These are the numbers we'll measure in our first pilots, starting from each store's own baseline.
F&I is where dealers get burned. Our agents are built to raise profit by working within each lender's guidelines and your store's policies, the same way on every deal.
Demo controls. They change this page only.
aimlark agents are built to work on top of Dealertrack, RouteOne, CUDL or whatever credit platform your store uses, never instead of them. They read, compare and draft; your managers approve. They're designed to work from the screens and exports your team already uses, so there's no new system to learn and nothing to rip out.
Today: screens and exports. Direct connections are on the roadmap.
A pilot starts in shadow mode: the agents work every deal next to your desk while nothing goes out, so you can judge them on your own deals before you rely on them.
We'll publish what our first pilot stores see here. No made-up numbers, no borrowed logos.
What dealers say, in their own words.
Coming after our first pilotsNumbers measured from each store's own baseline.
Coming after our first pilotsSanti saw it from the funding side, Irvin from the sales side. The managers knew how to work a deal. They just never had the time to work every approval against every program, on every deal.
So we're building agents that do the full math on every deal and hand your managers the answer, not the homework. Your people keep the relationships, the customers and the final call.
Santi and Irvin worked together in auto finance, Santi on the funding side and Irvin on the sales side. Ash leads our creative and knows how service businesses run on follow-through.
Ash was a Master Trainer at EOS Fitness in Utah, where he was named Top Trainer of the Year in 2025. He knows how service businesses run on relationships, scheduling and follow-through. At aimlark he focuses on industries that assume AI isn't for them, showing owners how agents can fit into the way they already work.
Santi co-founded aimlark because he believes AI agents can help businesses make more money than ever, and change the lives of the owners and the people who work for them. His background is in banking, finance and auto finance consulting, where he learned how deals really get done. He and the team also shoot and edit aimlark's films and commercials.
Irvin came up through auto finance on the sales side, owning a large book of dealer accounts and working daily with sales and finance managers. Where Santi saw how deals get bought and funded, Irvin saw how they get sold, and where the manual work slows every desk down. Together they understand the dealership deal from both ends, and Irvin makes sure what aimlark builds fits how managers actually work.
The same approach works anywhere skilled people lose hours to repetitive work. Tell us how your business runs, and we'll design agents around it and quote them after a free estimate.
Put in your store's volume and a what-if amount per deal. It's your math, not our promise. On a free estimate call we look at where that money actually is in your deals.
A what-if from your inputs: deals × share × amount × 12. Not a result or estimate from aimlark. Nothing you enter is sent.
A short call with a founder. No cost, no commitment.
Not yet. We're building it now and looking for our first dealers to shape it with us. Every agent on this page is marked Roadmap until it's running in a real store.
No. The agents take the grind: reading lender decisions, comparing structures, working callbacks, chasing stips. Your managers set the limits, handle what falls outside them, and present to customers.
That's the plan, inside the parameters you set. You choose the markup, term, advance, product and lender limits, and whether a manager approves each callback or the agents handle counters that leave the customer's deal unchanged and stay inside your limits. Anything outside your limits goes to a manager. With customers, your F&I manager always presents payment, rate and products.
Our agents are built to sit on top of Dealertrack, RouteOne, CUDL and other credit platforms, never instead of them. Today they work from the screens and exports your team already uses. Direct connections are on our roadmap.
By staying inside each lender's guidelines and your own policies, the same way on every deal. Products are always itemized and offered on one menu, and every decision is logged. Your compliance team and counsel own the rules; our agents aren't legal advice.
The ones your lenders already send your store: rate sheets, tier and advance guidelines, specials and stip requirements. The agents are built to keep them current and match each deal to the program that fits.
The finished deal: the best structure across every lender that approved, the rules it passes, what the store keeps, and the runner-up options. Managers approve and present. They don't redo the math.
Franchise stores and larger independents running real volume through Dealertrack, RouteOne or CUDL.
Yes. If you're not a dealership and want an agent tailored to your business, we'll build one. Tell us about it on a free estimate call and we'll quote it.
Every system is built around your store, so we quote after a free estimate. You decide from there.
Santiago Quiroz, Ashton Hamm and Irvin Valencia, the founders of aimlark. You talk to us directly.
Let's look at where your deals stall and what agents could take off the managers you already have.