Set the goals. Let an AI agent negotiate them.
Describe the deal in plain language. The AI drafts the full scenario — objectives, guardrails and trade-offs — then negotiates with the supplier on your behalf, showing every step in a transparent reasoning log.
A single price field, one dimension, no follow-up.
Draft it in words. Let the agent run it.
Describe the deal once. The AI turns it into a structured scenario — objectives with priorities, guardrails it can't cross, behavior policies, and the trade-offs it's allowed to offer — then negotiates against all of them at once, with the buyer keeping oversight and the final say.
What procurement teams get out of it.
A negotiator that runs your playbook on every request — and stays inside the lines every time.
More than price on every deal.
The agent works several objectives at once — price, payment terms, volume discount, price hold — instead of leaving value on the table on everything but the headline number.
Drafted from a prompt
Type the deal in plain language — the AI fills in scope, objectives and trade-offs for you to review.
Guardrails it can't cross
Hard limits and behavior policies — never reveal the limit, no pressure tactics — are enforced automatically.
Trades value, not principles
It knows what it can give — longer term, more volume, faster payment — and what must never move.
Every move explained
A live reasoning log shows why the agent made each offer — no black box, fully reviewable.
How the agent works under the hood.
The buyer defines the goals and the limits; the agent executes them. Everything is transparent, auditable, and stays under human control.
Draft with AI
A plain-language prompt becomes a structured scenario — scope, objectives and trade-offs, filled in ready for review.
Multi-objective goals
Price, payment terms, volume discount and price hold — each with its own current, target and priority weight.
Guardrails & policies
Hard limits plus behavior rules — don't reveal the secret limit, no pressure tactics, no internal context shared.
Trade-off engine
Defines what the agent may offer in exchange — volume, longer contract, forecast commitment, faster payment — and what stays fixed.
Transparent reasoning log
Every offer comes with the why — which lever it used, which guardrail it respected — logged step by step.
Dry run & human control
Rehearse the whole negotiation before it goes live. Approval and awarding always stay in human hands.
4 steps from a prompt to a closed deal.
Draft with AI
Step 1 of 4
Draft with AI
Describe the deal in a sentence. The AI fills in the scope — product, volume, window — and proposes the rest for you to review and adjust.
Set objectives & guardrails
Confirm each objective's target and priority, and the hard limits the agent can never cross — including a secret limit it must not disclose and behavior policies it must follow.
Define the trade-offs
Tell the agent what it's allowed to give in exchange for a better price — and what must stay untouched. It only ever trades within these bounds.
Agent ↔ Supplier
Round 6 · closed
Agent negotiates — you award
Rehearse it as a dry run, then let it go live. The agent works the objectives round by round inside your guardrails, logs every step, and hands the result back for you to review and award.
Your best negotiator's playbook — running on every request.
prompt
objective
guardrails
logged
“This is a real game changer — everybody talks about automation but you really deliver it at scale for us.”
Procurement teams in chemicals, pharma, and coatings
build with PINPOOLS.
Bot Negotiation questions, answered.
Does the agent make the awarding decision?
No. The agent negotiates within your objectives and guardrails, then hands the outcome to the buyer for review and awarding. Approval always stays with a human.
Can I test it before it talks to a supplier?
Yes. Run the whole negotiation as a dry run first — the agent plays through the rounds against a simulated counterpart so you can check its behavior, then switch it live when you're comfortable.
How do I know it won't cross a line?
You set hard limits and behavior policies up front — a secret limit it must never reveal, no pressure tactics, no sharing of internal context. The reasoning log shows exactly which guardrail applied at each step.
Is Bot Negotiation available today?
Bot Negotiation is in Beta — available to selected customers under early-access. Talk to us to join the next cohort.
See the agent negotiate your goals — on your own deal.
20 minutes with our team — draft a scenario and run a live negotiation on your categories.
Trusted by industry leaders · Enterprise security · Onboarding in days