Market shortlist
Hide duplicate and low-fit notices, then surface opportunities that match your profile, regions, capacity, and timing.
Aeryl
Procurement intelligence
with altitude
Find relevant tenders faster, understand buyer activity, and monitor the markets that matter.
Thesis
Not an AI wrapper.
Proprietary procurement models.
Aeryl reads tenders, awards, lots, criteria, suppliers, buyers, prices, and relationships as one procurement graph, then turns that graph into decisions that growth and bid teams can inspect.
INPUT / ONE TENDER RECORD ENTERING AERYL tender_id: EU-2026-01844 title: municipal fleet renewal / electric vans buyer: city transport office procedure: open tender deadline: 2026-06-18 cpv: 34144900 electric vehicles lots: L01 42 vehicles est. EUR 1.8M - 2.4M L02 charging service est. EUR 320K - 480K award_criteria: price 60 | delivery 20 | service 20 linked_memory: 12 buyer awards | 2 incumbents | 41 suppliers raw_notice >>> normalize >>> resolve_entities >>> graph_link graph_link >>> feature_store >>> AERYL model stack
Product
From tender noise
to bid decisions.
Aeryl is not a tender inbox. It is a qualification layer that turns notices, buyer memory, lot structure, and model evidence into bid decisions your team can act on.
Buyer memory
Read renewal cadence, incumbent gravity, comparable awards, lot patterns, and buyer preferences before committing effort.
Bid decision
Review AERYL-ALIGN, AERYL-PRICE, AERYL-COMPETE, and AERYL-TRACE as one inspectable recommendation.
Model systems
Proprietary models,
trained on procurement.
Three procurement-native systems read a small, controlled feature store. The website shows the shape of the models without exposing training recipes, weights, or scoring internals.
ArchitectureAERYL-ALIGN
AERYL-ALIGN / RELEVANCE RANKING MODEL feature_store buyer.fit[441] sector.vector[034] lot.complexity[17] deadline.window[22] profile.match[**] capacity.window[**] candidate_set notices[open] lots[structured] buyer_memory[linked] low_fit -> suppressed duplicates -> folded rank pass profile + sector + timing + buyer fit >>> shortlist score + relevance reason output: opportunity rank, fit band, reason slots eval: nDCG@20 0.506 | MRR 0.460 | recall@50 0.868
Workflow
From signal
to action.
Aeryl is built for the daily bid loop: define the company, rank the market, inspect the evidence, decide what to pursue, and keep watch when the tender moves.
Private preview
For teams that need earlier tender discovery, buyer intelligence, model evidence, and change monitoring in one operating rhythm.