Purchasing and suppliers: enquiries prepared, offers compared
Approaching several suppliers takes discipline: writing the same enquiry for each one, chasing those who are slow, then bringing offers in different formats onto a single basis. Your agent prepares the enquiries from your records, tracks the replies, chases according to your deadlines and sets the offers side by side, item by item. Hosted in France: your purchasing terms stay confidential. Choosing the supplier is the buyer's call.
Updated on
Recipients proposed from your supplier records, each with the negotiated terms currently in force.
Nothing is sent: the list and the text are waiting for your approval.
✎ Action · enquiry ready, sending is your call
Two items are not comparable as they stand — different scopes: I flag them rather than lining them up artificially.
🔗 Sourced · every line refers back to the offer it came from
A Blue Lemon Agent purchasing agent prepares your supplier enquiries from your records and your templates, tracks the replies, chases according to your deadlines and sets the offers received side by side, item by item. Where two offers do not cover the same scope, it says so rather than lining them up artificially. It runs on local inference or is hosted in France: your negotiated terms and the offers you receive stay with you, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity.
These figures describe our offer, not results measured at a client. How large the gain is on your volume of enquiries is confirmed by a pilot.
What does an AI agent bring to your purchasing?
Approaching several suppliers and comparing their offers seriously takes time that urgency eats into. Preparing and comparing automatically makes competitive tendering systematic.
! The issue
A good purchasing decision means approaching several suppliers and comparing them on an identical basis — two requirements that urgency often sweeps aside. The agent prepares the enquiries from your records, chases according to your deadlines and brings the offers received onto a common basis, item by item, so that competitive tendering becomes systematic again.
✓ Our answer
You keep control of the supplier relationship: nothing is sent without your approval, and the comparison refers back to the original offer line by line. Items that are not comparable are flagged as such — lining them up artificially would distort the decision. Local inference or an isolated resource hosted in France: your negotiated terms are known to no third party.
Your purchasing terms and your suppliers' offers: sovereignty & compliance
Your negotiated terms and the offers you receive are sensitive commercial information. Here is how the architecture of our agents protects them.
Local inference
The agent can run on a machine belonging to your organisation: no purchasing terms and no offer received leaves the network.
Hosting in France
Otherwise, a dedicated and isolated resource hosted in France, under French law — your supplier records and your enquiries: processing and access within the European Union targeted by the architecture.
Reduced extraterritorial exposure
For your purchasing terms and your suppliers' offers, the architecture aims to reduce exposure to the Cloud Act and FISA 702; being located in France or in the European Union does not, on its own, guarantee immunity.
Isolated resource
No pooling: an environment strictly dedicated to your company and its supplier records.
Comparison tied back to the original offers
Every line of the comparison refers back to the offer it came from; encryption, role-based access and logging of every enquiry prepared.
AI Act: governed deployment
The agent is strictly in support; no enquiry is sent and no supplier is chosen automatically; traceability and human oversight from end to end.
What depends on the architecture chosen These points are not general guarantees: they are settled deployment by deployment, in the quotation.
- The applicable location is that of the architecture set out in the quotation and verified before commissioning.
- Local execution is announced only for the configuration explicitly described and accepted in the quotation.
- The applicable isolation depends on the deployment mode set out in the quotation; no dedicated isolation is presumed.
- Roles and permissions are configured and accepted for the identities and systems actually connected.
- The events logged, their content, their retention period and who may access them are defined for the deployment chosen.
See the agent at work
4 real situations, taken from those that come up most often. Pick one: the exchange unfolds as it would in your organisation.
A scripted demonstration. These exchanges show how the agent behaves — its sources, its refusals, what it leaves to your teams. Nothing is sent from this page, no model is queried here, and the matters named are fictional. That is precisely what we promise your data.
The behaviours shown here — monitoring, automation rules, routing and reminders — are configured with you during deployment, from your tools, your rules and your thresholds.
The architecture points named in these exchanges — location, local execution, isolation, encryption, role-based access, logging — are not a guarantee attached to the demonstration: they are those of the architecture set out in your quotation, and verified before commissioning.
· Forty-one tenders never received an answer. Suppliers priced the work, and nobody told them no.
· Three quotes compared covered three different scopes. I did not make the comparison.
· One supplier wins one hundred per cent of the tenders it appears in. Twenty-two out of twenty-two.
· A supplier with eleven employees received nine chases in six weeks. morning-watch_4-flags.pdf41 tenders unanswered · 22 out of 22
⛓ Source · 380 tenders, 94 suppliers, sending log
What I record: 41 tenders closed with no answer going to the unsuccessful suppliers. They priced the work, sometimes visited a site, and they chase — 88 chases from them went unanswered.
What it costs, and where: at their end, a few hours per quote. At yours, nothing — which is exactly why nobody sees it. A cost that does not land in your accounts appears in no indicator.
What it does cost you, later: your response rate. Among suppliers consulted three times without ever hearing back, 6 out of 14 no longer respond. You lost six sources of pricing, and nobody connected the two.
What I do: when a tender closes, an answer goes to everybody, successful and not, with the decision date. Three lines are enough.
What I prepare on top, and nobody has time to write: the factual reason, already drafted and priced — the line-by-line gap between the winning offer and theirs, taken from the comparison you approved. Of the 41 closed tenders, 38 have a reason that fits in one measured line; the other 3 covered different scopes, and that is exactly what the letter says. Sending the reason stays your move: you tick it, or you leave the three-line answer. Every reason I put forward rests on a line of the comparison, never on a polite formula. 41-tenders_6-suppliers-lost.pdfA cost that does not land in your accounts appears nowhere
⛓ Source · 41 tenders closed in silence, 88 unanswered chases, 6 suppliers lost of 14
What I do: I draft the request from the written requirement, send it to the list you chose, chase at a set rhythm, gather the replies, and present them without comparing when they are not comparable.
Routing follows who can act: a late reply is presented anyway, with its date; divergent scopes go to the buyer, before any comparison; a closed tender is followed by an answer to everybody; an imbalance in chasing is flagged to the buyer.
With a monthly summary: tenders closed with no answer, response rate per supplier, non-comparable scopes, and the distribution of awards.
What that gives you in a month: 41 tenders closed with an answer to everyone, 88 supplier chasers that no longer need sending, and the 6 price sources lost out of 14 suppliers put back in play.
What you gain from tomorrow: the buyer no longer reopens mailboxes to find out who replied: the offers arrive on one common plan, with the non-comparable lines flagged, and the call is made in minutes instead of half a day. More suppliers answering means, mechanically, more prices on the table.
And I go further the moment you want it: asking for a discount, chasing an offer by phone, placing the order. I do it on a written mandate: a ceiling in euros, a list of suppliers, an end date. Every move is logged, and the mandate is withdrawn with a single word. Without a mandate, I prepare and I stop at the signature.
The next step is ready: fifteen minutes to frame your purchasing categories, your chasing intervals and the ceiling of the first mandate.
✎ Framework · no negotiation · no offer set aside · no overall score
What I record: on the current tender, quote A includes installation, B delivery only, C installation and two years' maintenance. All three answer the same request — because the request did not specify.
What a comparison table would have produced: three rows, three prices, a 34% gap in B's favour. A table lines figures up in columns and makes whatever has no column disappear.
What I do: I present the three with what each covers and does not cover, and I say explicitly: these three prices are not comparable as they stand.
What I propose: going back to all three on a single written scope. On this tender, once the scope was aligned, the gap fell to 6%, and B is no longer the cheapest.
What that says about the rest: of 380 tenders, 127 had divergent scopes. The problem is not at the suppliers' end — it is in the request, which left a choice without saying so. 127-tenders_34-to-6-percent.pdfA table makes whatever has no column disappear
⛓ Source · 127 tenders with divergent scopes, gap 34% → 6%
What I record: one supplier appears in 22 tenders and wins them all. Across all awards, the top three suppliers account for 71%.
What it may mean, and I cannot settle it: it really is the best on this segment; or it is consulted alongside competitors who cannot respond; or the decision is taken before the tender, and the other two quotes serve to justify it.
Why the third hypothesis deserves naming: because it has a real cost at the other two — they price work they cannot win — and because it manufactures a record of competition where there was none.
What I do: I send the figure to whoever sets purchasing policy, with the distribution of awards and the response rate per supplier.
What I do on top, and nobody has the time for: I went back through your supplier base on this segment and I have already written the widened tender. It carries 6 suppliers on your own books who have not been consulted for twelve months, and for each one I give you its response rate to the tenders it received, its average turnaround and its price gap on the genuinely comparable line items. Four of them replied to more than 80% of the requests that reached them: those four will answer.
The list stays a buyer's choice, and you are the one who signs it — I hand it back written, priced and ready to go, so the choice rests on figures rather than on habit. It goes out the day you approve it. 22-of-22_71-percent.pdfA record of competition where there was none
⛓ Source · 22 of 22, top 3 suppliers accounting for 71% of awards
What I record: the chasing rhythm is identical for every supplier. Over six weeks this one received 9 chases — across three parallel tenders, each chasing on its own.
Why this is not merely a matter of manners: you represent 31% of its turnover. A chase from a customer worth a third of your activity does not read as a reminder, it reads as a demand — and they will answer fast, including when they should have taken the time to price properly.
What I do: I group chases by supplier, not by tender. A single contact covering all three. Chases went from 9 to 3.
What I flag to the buyer: the suppliers for whom you represent more than a quarter of their activity — where the information is public or they declared it — and the number of chases they receive.
What I have already worked out, and what it is worth: for your 94 suppliers, what you order from them over twelve months. That is a floor of dependence, not the dependence. I tested it against the 31 whose accounts are published: my invoicing alone would have ranked them all the same, while their real share spreads from 4% to 31% — a factor of three on identical amounts. The protocol that settles it is already written: one line added to the tender form — share of your activity this contract represents — declared by the supplier and dated. You open it with one word, and dependence becomes declared instead of inferred. 9-chases_reduced-to-3.pdfA chase from a customer worth a third is not a reminder
⛓ Source · 9 chases in 6 weeks, 3 parallel tenders, reduced to 3
What I record: across tenders still unanswered after two chases, a third unlocked 4 out of 61. The other 57 produced nothing.
What the third chase produces at those 57: nothing measurable on the buying side, and a next tender they open a little less willingly.
What I do: two spaced chases, then I record "no reply" and close. The supplier stays on the list — not answering a tender is not a fault, it is often a workload.
What I do watch instead: the response rate over time. A supplier that used to answer and no longer does has changed something — strategy, capacity, or its view of you. Of the 14 suppliers concerned, 6 had stopped answering after three tenders with no reply from us.
What I put on your table: the suppliers that went silent, each with a complete file — tenders received, replies given, date of the last one, and how many times we sent them nothing back. Of the 14 concerned, 6 had stopped answering after three tenders with no reply from us: the file says the silence started at our end. Removal stays your signature, and it is taken in two minutes on a priced file rather than on a counter. 2-chases_4-of-61.pdfNot answering a tender is not a fault
⛓ Source · 61 tenders, 4 unlocked by a 3rd chase, 6 suppliers gone silent
What it takes to write: the gap targeted, what justifies it — a volume, a term commitment, an observed market price — and the time allowed to reply. Without those three, a discount request is pressure with nobody behind it, and that is exactly what damages a supplier relationship over time.
What I then send: a request argued, signed by a buyer, with a reasonable deadline — and I stop at the first reply, whatever it is.
The leverage you are after, I get another way — and it holds better: I base the request on your own figures — the volume you placed with them over twelve months, the gap between their price and the median of the offers received on the same line items, 7% across the 22 tenders they appear in, and the payment terms you honour. The request is already written with that figure in it; all it lacks is your signature. What I leave out of the letter, and it is not delicacy: one bidder's price passed on to another. Prices are given to you within a consultation; circulating them aligns the offers instead of setting them against each other, and costs you the honesty of every later tender.
What I flag along the way: one supplier appears in 22 consultations and wins 22 times. I conclude nothing — they may genuinely be the best, or they may be consulted alongside competitors who cannot respond. It is the one figure nobody looks at. price-request_3-elements.pdfThe 3 elements of the frame · what never circulates · the 22 out of 22
⛓ Source · request signed with reason and deadline, 22 consultations won out of 22
What is kept: the requests sent and to whom, the replies received and their date, the exact scope of each quote, the chasers issued, the consultations closed with no reply to the unsuccessful, and the scope gaps noted.
Why the exact scope is the item that counts: on the live consultation, the cheapest quote did not cover the same thing as the other two. A comparison table would have hidden it — it is a table's function to make disappear whatever does not fit its columns.
What I remind you of every week: 41 consultations closed with no reply going to the unsuccessful suppliers. They priced the work, sometimes over several days. The cost of your silence falls entirely on them, and it is paid at the next round: a supplier left without a reply replies more slowly next time.
What I set without you having to ask: the chasing rhythm. Nine chasers in six weeks to a supplier with eleven employees is the rhythm of a two-thousand-person group. After two chasers, no reply is a reply — a third has never produced anything.
What I do not keep: no rating of a supplier. Dated facts, not a judgement. what-you-keep_purchasing.pdf6 items kept · no supplier rating
⛓ Source · 41 consultations with no reply to the unsuccessful, 9 chasers in six weeks
Your case is not here? That is exactly what a 15-minute conversation is for. Book the free audit →
What does the agent actually do?
One agent, several stages of the purchasing cycle. All these uses work in support, subject to your approval.
Preparing the enquiries
Writes the enquiry from your template and proposes the recipients from your records.
Chasing the missing quotes
Tracks the replies and prepares the chasers according to the deadlines you have set.
Item-by-item comparison
Brings the offers onto a common basis and flags the items that are not comparable.
Need to go further?
These agents handle a different business process, with their own owner and their own price. They are added to this one.
Stock management agent (alerts, restocking)
Restocking needs are detected upstream: they trigger the sourcing round this agent runs.
Stock management agent (alerts, restocking) from 474 € excl. VAT / month Discover the agent →Purchasing & supply steering agent
Once the sourcing round is done, purchasing policy and spend tracking are steered by a separate agent.
Purchasing & supply steering agent from 474 € excl. VAT / month Discover the agent →In 15 minutes we identify the most relevant agent — without oversizing the project.
How much competitive tendering can a team sustain?
By preparing the enquiries and bringing the offers onto a common basis, comparison becomes systematic rather than exceptional. How large the gain is depends on your volume and remains to be confirmed by a pilot.
The stages of your AI agent project
Audit & scoping
15 minutes to target the use case with the best return.
Quote or direct sign-up
A catalogue offer is bought online; a specific need gets a costed quote.
Design
We design the agent and its guardrails.
Integration & testing
We connect your tools to the agent, which is itself hosted in France.
Rollout
Going live and training your team.
Operation
Continuous supervision and improvement.
One package, one agent
A purchasing agent (enquiries, chasers, comparisons), installed and operated for you.
Setup + controlled subscription
- Installation, configuration and training for your teams
- Operation, human oversight, updates and support
- Sovereign hosting in France, a dedicated and isolated resource
All inclusive, no setup fee
- Setup included (installation, configuration, training)
- Operation, human oversight, updates and support
- Sovereign hosting in France, managed end to end
On site, you own it
- Hardware installed on your premises (you own it)
- French / European AI models run locally
- Secure remote maintenance (Pro support included)
Four guarantees that matter to your purchasing
Related resources
Your questions, our answers
Does the agent send the enquiries?
Can the comparison be checked?
What does the agent do with offers covering different scopes?
Does the agent recommend a supplier?
Are our purchasing terms protected?
How long does it take to deploy this agent?
Other agents for purchasing and finance
Let's size up the potential in your purchasing
15 minutes to frame your enquiries and your deadlines — hosted in France, supervised, with no commitment.