Confidential proposal · Prepared for Yalago

Supplier Performance & Intelligent Acceleration Program

Helping Yalago improve supplier response times, increase search completion within the 3-second SLA, reduce infrastructure costs and build a future-ready hotel distribution platform.

Review Proposal Schedule Workshop

Peak search volume

500K/min

Internal search

<40ms

Current SLA

Throttled requests

~10%

Executive Summary

Yalago's internal platform already delivers excellent performance.

Internally managed inventory and PUSH suppliers respond in under 40 milliseconds. Based on the information shared, the primary opportunity for further optimisation lies in downstream PULL supplier integrations, where higher response times become the main constraint on end-to-end search performance. When supplier latency exceeds the target 3-second SLA, searches may be completed without all available options, impacting search completeness, increasing infrastructure load, and limiting the platform's ability to consistently deliver the best available inventory.

Platform signals at a glance

Operating metrics shaping the opportunity

280K–500K

Searches per minute

Peak query volume across the distribution platform.

3 s

Search SLA

Hard ceiling after which results are terminated.

~10%

Throttled / rejected

Share of requests lost or downgraded at peak.

Our proposal

Cognirock proposes a Supplier Performance & Intelligent Acceleration Program designed to optimise supplier performance, improve search completion and prepare the platform for future growth.

Phase 101

Assessment & Diagnosis

Analyse the current platform performance to identify the main bottlenecks, evaluate PULL supplier behaviour and define a data-driven optimisation roadmap.

Per-supplier performance benchmark
Identification of quick wins
Latency and response-time analysis
Business impact assessment
Prioritised improvement roadmap

Phase 202

Proof of Concept (PoC)

Design and implement a Proof of Concept to validate, in a real environment, the improvements identified during the diagnosis and demonstrate their impact before a full rollout.

Integration of selected suppliers
Validation of the acceleration strategy
Before vs. after performance comparison
Measurement of results and KPIs

Phase 303

Development & Rollout

Build and deploy the solution validated during the PoC into production, integrating it with Yalago's architecture and ensuring a controlled transition.

Development of the solution
Deployment to production
Integration with the defined suppliers
Performance monitoring and validation

Phase 404

Follow-up & Optimisation

For the 6 months following go-live, we run continuous follow-up to measure performance, optimise operation and support Yalago's team through adoption. At the end of the period, Yalago can opt into an ongoing support and evolution service covering maintenance, new features and continuous optimisation.

KPI and performance tracking
Continuous optimisation of the solution
Periodic review meetings
Improvement recommendations
Optional support and evolutionary maintenance service

Expected outcome

A platform capable of serving more searches within the SLA, reducing supplier API traffic and infrastructure costs, improving resilience, and ultimately increasing booking conversion while supporting Yalago's continued global growth.

Current Architecture

Parallel supplier fan-out — bounded by the slowest path.

Every search fans out to contracted inventory, PUSH and PULL suppliers in parallel. PUSH and contracted responses are fast; PULL suppliers introduce highly variable latency and frequent timeouts.

Contracted + PUSH

< 40 ms · deterministic latency · in-house indexes.

PULL suppliers

Variable latency · timeouts · throttling at peak.

Outcome today

Searches terminated at 3 s → incomplete results → lost upside.

The Opportunity

Four leverage points where engineering meets commercial value.

Opportunity

Reduce downstream latency

Insulate Yalago from slow PULL suppliers using intelligent acceleration and adaptive freshness — not blunt caching.

More results inside the 3 s SLA.

Opportunity

Increase completed searches

Serve frequently requested inventory from an acceleration layer while preserving real-time accuracy.

Higher conversion potential.

Opportunity

Reduce supplier traffic

Lower the number of outbound API calls during peaks — directly addressing throttling and rejection rates.

Lower throttling and cost.

Opportunity

Improve platform scalability

Decouple search growth from supplier traffic growth and from raw infrastructure scaling.

Headroom for next-stage volume.

Program Roadmap

Four phases — from diagnosis to a continuously optimised platform.

Each phase is independently valuable and builds on the previous one. Below is the expected timeline from kick-off to the end of the follow-up period.

Expected timeline

≈ 12 months end-to-end

013–5 weeks

Weeks 1–5

024–6 weeks

Weeks 6–11

033–4 months

Months 4–6

046 months

Months 7–12

Phase 1

Assessment & Diagnosis

Weeks 1–53–5 weeks

Analyse the current platform performance to identify the main bottlenecks, evaluate PULL supplier behaviour and define a data-driven optimisation roadmap.

Per-supplier performance benchmark
Identification of quick wins
Latency and response-time analysis
Business impact assessment
Prioritised improvement roadmap

Phase 2

Proof of Concept (PoC)

Weeks 6–114–6 weeks

Design and implement a Proof of Concept to validate, in a real environment, the improvements identified during the diagnosis and demonstrate their impact before a full rollout.

Integration of selected suppliers
Validation of the acceleration strategy
Before vs. after performance comparison
Measurement of results and KPIs

Phase 3

Development & Rollout

Months 4–63–4 months

Build and deploy the solution validated during the PoC into production, integrating it with Yalago's architecture and ensuring a controlled transition.

Development of the solution
Deployment to production
Integration with the defined suppliers
Performance monitoring and validation

Phase 4

Follow-up & Optimisation

Months 7–126 months

Continuous follow-up during the 6 months after go-live to measure performance, optimise operation and support Yalago's team through adoption. Optional ongoing support and evolution service afterwards.

KPI and performance tracking
Continuous optimisation of the solution
Periodic review meetings
Improvement recommendations
Optional support and evolutionary maintenance service

Intelligent Refresh Strategy

Freshness is adaptive.

Instead of one cache lifetime for everything, refresh decisions are derived continuously from product, market and supplier signals — keeping accuracy high and supplier traffic low.

Lead time

Closer arrival → fresher data.

Popularity

Hot products refresh more often.

Seasonality

Peaks tighten refresh windows.

Destination demand

Hot destinations push priority.

Supplier health

Healthier suppliers refresh harder.

CheckRate mismatches

Discrepancies → immediate refresh.

Supplier Intelligence

A live operating picture of every supplier on the distribution graph.

Continuous monitoring of response times, error rates, availability consistency, supplier health and refresh efficiency — feeding the acceleration layer in real time.

Supplier Health Board

last 5 min · auto refresh

Supplier	Avg RT	Health	Hit ratio
Supplier Alpha	38 ms	98	92%
Supplier Bravo	220 ms	86	71%
DMC Cobalt	1,840 ms	64	38%
DMC Sierra	2,640 ms	41	22%
Supplier Quartz	95 ms	94	88%
Specialist Onyx	2,980 ms	32	14%

Avg response time

↓ 64%

◆

Cache hit ratio

78%

◆

Timeout rate

↓ 71%

◆

Refresh efficiency

92%

◆

Business Impact

Engineering outcomes that compound into commercial outcomes.

Supplier performance

Faster searches

More results

Higher conversion

More revenue

Higher

Search completion within SLA

Lower

Supplier API calls during peak

Lower

Infrastructure load on ARI compute

Lower

Throttling and rejection rates

Better

SLA compliance under traffic spikes

Higher

Conversion from richer result sets

Greater

Platform resilience and headroom

Better

Customer experience end-to-end

Why Cognirock

Not a connectivity platform. Not a cache. A partner for supplier performance at scale.

Deep travel expertise

Hotel distribution, connectivity, API optimisation and marketplace technology — built specifically for this industry, not retrofitted from generic infra.

Hotel distribution
Connectivity protocols
API optimisation
Marketplace technology

Engineering excellence

We design systems for the constraints that matter at Yalago's scale — latency tails, supplier behaviour and operational headroom.

Performance optimisation
Supplier orchestration
Distributed architecture
Scalable infrastructure

Proven technology

Bundleport already operates as a hotel connectivity marketplace — the same foundation powers the supplier intelligence and acceleration layer.

Bundleport marketplace
Supplier intelligence
Acceleration platform
Production-grade tooling

Investment

Phased commitment aligned to the program roadmap.

A dedicated team of 3–4 engineers running the full program end-to-end — from diagnosis to production rollout and 6 months of follow-up.

Phase 1

Assessment & Diagnosis

€9,000

3–5 weeks

Per-supplier performance benchmark
Identification of quick wins
Latency and response-time analysis
Business impact assessment
Prioritised improvement roadmap

Phase 2

Proof of Concept (PoC)

€15,000

4–6 weeks

Integration of selected suppliers
Validation of the acceleration strategy
Before vs. after performance comparison
Measurement of results and KPIs

Core build

Phase 3

Development & Rollout

€30,000

3–4 months

Development of the solution
Deployment to production
Integration with the defined suppliers
Performance monitoring and validation

Phase 4

Follow-up & Optimisation

€6,000

6 months

KPI and performance tracking
Continuous optimisation of the solution
Periodic review meetings
Improvement recommendations
Optional evolutionary maintenance service

Total investment across the four phases is €60,000 — final price with a 10% discount already applied, covering a dedicated team of 3–4 people running the program for close to a year.

Billing plan

Payments tied to program milestones

Invoicing is split into three milestones aligned with the delivery of each stage of the program.

30%€18,000

01At project kick-off

40%€24,000

02On completion of Phase 2 (PoC)

20%€12,000

03On completion of Phase 3

10%€6,000

04At the end of Phase 4

Next step

Let's build the next-generation supplier performance platform.

From reducing latency today to creating the most intelligent hotel supplier acceleration layer in the industry.

Schedule Technical Workshop Contact Cognirock