Advice, delivery and operations
from one accountable team.
We advise on strategy, design the architecture, deliver the platforms and then run them to bank-grade SLAs for banks, government entities and large enterprises across Saudi Arabia, Jordan and the wider Middle East.
Strategy, architecture and regulatory readiness
For CIOs, COOs and heads of digital who must commit to a platform, a vendor or a regulatory deadline and defend it to the board.
Platform and vendor decisions are expensive to reverse, and the board, internal audit and the regulator will review them. Our advice comes from operators who have owned these decisions inside the Saudi financial sector.
- Integration and platform strategy, target architecture
- Regulatory readiness: SAMA frameworks, Open Banking, PDPL, NCA ECC
- RFP authoring, vendor evaluation and negotiation support
- Design authority and architecture review as a service
What you get
- Architecture blueprint: current state, target state and the decisions behind them
- Integration inventory and dependency map
- Regulatory gap assessment for the scope in question
- Target operating model: ownership, teams and design authority
- Costed, sequenced roadmap with a board summary
How it typically runs
- Frame Week 1Agree the decision, the scope and who signs off.
- Discover Weeks 1 to 3Interviews, documents and an inventory of systems and interfaces.
- Assess Weeks 2 to 5Options, costs and risks, tested with your architects.
- Recommend Final weekReadout to the sponsor and the written pack.
Products it draws on
- ConsentHubOpen banking platform
- PayConnectPayment integration suite
- KYCLinkIdentity and KYC gateway
Typical scenario
A bank faces an open banking milestone with two competing vendor proposals. Over about four weeks we map what the framework requires against what the bank runs today, score both proposals against the gap and give the steering committee one costed recommendation.
Integration platform and API delivery
For banks and government entities launching products, payment rails or e-services that depend on the integration layer.
Every new channel, rail and partner adds connections, and point-to-point fixes slow each launch. We deliver the gateways, service buses, event streaming and payment and government connectivity that must work at 2 a.m. on settlement day, with idempotency, circuit breakers and observability from the first release.
- API platforms and gateways (API Connect / DataPower-class)
- ESB and iPaaS delivery: IBM ACE / CP4I, MQ, Kafka
- ISO 20022 payments: SARIE, IPS, SADAD and bulk
- Government APIs: Nafath, Yakeen, National Address, GOSI
What you get
- Solution architecture and interface specifications (OpenAPI, AsyncAPI)
- Flows, APIs and adapters in your source control, with CI/CD pipelines
- Security design: mTLS, JWS signing, replay defence and FAPI token profiles
- Automated test packs and performance evidence for every release
- Runbooks, monitoring dashboards and a handover pack
How it typically runs
- Mobilize Weeks 1 to 2Team, environments, backlog and design authority in place.
- First increment Weeks 3 to 6One thin slice, end to end, released to production.
- Further increments Every 2 to 4 weeksMore flows and channels, each with test evidence.
- Hand over 4 to 6 weeksHypercare, runbooks and knowledge transfer to your team.
Products it draws on
- PayConnectPayment integration suite
- KYCLinkIdentity and KYC gateway
- VerifyFlowEarly accessVerified onboarding
- EnforceDeskEnforcement execution for Tanfeeth
- NotifyHubNotification hub
Typical scenario
A government entity launches an e-service that needs each applicant's identity, address and employment record. We connect Nafath sign-in with Yakeen, National Address and GOSI checks through one integration layer, with consent recorded and every call traced.
Legacy integration modernization and migration
For CIOs with an integration estate nobody can fully explain, on platforms nearing the end of vendor support.
We reverse-engineer undocumented systems into flows, queues, data models and business rules, then migrate in waves using strangler patterns. Each wave is proven against recorded production traffic before the old path is switched off.
- Reverse-engineering of undocumented integration estates
- Strangler-pattern migrations with recorded-traffic replay
- Containerization to OpenShift / Kubernetes, GitOps and CI/CD
What you get
- Estate documentation: HLD, LLD and API specifications
- Rationalization register with a retire, consolidate or rebuild decision per service
- Migration plan in waves, with a rollback for each wave
- Replay evidence for every cutover
- Decommissioning plan for the legacy platform
How it typically runs
- Discover 3 to 6 weeksInventory and document the estate.
- Decide 2 to 3 weeksRationalization and the wave plan, agreed with owners.
- Migrate Waves of 6 to 10 weeksBuild, replay-test and cut over one domain at a time.
- Retire After the last waveSwitch off the old platform once reconciliation is clean.
Products it draws on
- DataShiftData migration platform
- OpsCommandOperations command center
Typical scenario
An integration platform carries several hundred services and is close to the end of vendor support. Discovery shows which services can be retired. The rest move in domain waves, each replayed against recorded production messages, while the old platform runs alongside until the last wave.
24/7 operation of your integration platforms
For COOs and heads of IT operations who need integration platforms run to an SLA without building a 24/7 team.
Integration failures often start quietly, with a stuck queue, a certificate close to expiry or a batch running late. We operate your integration layer 24/7 with NOC-grade monitoring and the SLAs your regulator and customers expect, reported monthly at executive level.
- 24/7 monitoring, incident and problem management
- Observability: distributed tracing, business activity monitoring
- Release management, capacity planning and DR drills
- Certificate, key and secrets lifecycle management
What you get
- Agreed service levels and error budgets for each platform, with SLA credits
- Operational runbooks and on-call cover for nights and weekends
- Monthly executive report: availability, incidents, risks and capacity
- A tested DR plan with drill evidence
- An exit plan, kept current, for handing the service back
How it typically runs
- Due diligence 2 to 4 weeksReview platforms, monitoring, open incidents and known risks.
- Transition 4 to 8 weeksRunbooks, access and monitoring, with shadow support beside your team.
- Steady state Ongoing24/7 operation, monthly reviews, quarterly improvement plans.
- Exit When you chooseHandback to your team or a new provider, using the exit plan.
Products it draws on
- OpsCommandOperations command center
- MailTriageAI triage for operational email
Typical scenario
A bank's integration team is strong on delivery but stretched at night and over weekends. We take first- and second-line support, and its engineers keep third line for change. Batches and certificate renewals are tracked in OpsCommand, and the COO receives one monthly service report.
AI automation of back-office work, inside your environment
For operations leaders with high-volume manual work in mailboxes, reconciliations or registers, and a compliance team that must approve any use of AI.
Many enterprise AI projects stall at proof of concept because they were never designed for the security review. We start with that review: agents run inside your boundary, on your data, with human approval on consequential actions, and are measured against a baseline your CFO will accept.
- AI agents for operations: triage, reconciliation, registers
- Document, email and workflow automation
- AI governance and PDPL-aware data handling
What you get
- A baseline of the current process: volumes, handling time and error rates
- A working agent on one workflow, in production-shadow
- Governance pack: data flows, access, approval points and audit logging
- Day-30 results against the baseline, with the economics
- A production rollout plan if the case holds
How it typically runs
- Baseline Week 1Measure the workflow as it runs today.
- Set up Weeks 1 to 2The agent is connected to the workflow inside your environment.
- Shadow Weeks 2 to 4The agent works beside your staff and its output is compared with theirs.
- Decide Day 30Go or no-go on measured economics.
Products it draws on
- MailTriageAI triage for operational email
- AssetSenseAssets register automation
Typical scenario
An operations team sorts a busy shared mailbox by hand. In a 30-day pilot, MailTriage classifies and routes the same mail in shadow mode. At day 30 the team compares accuracy and handling time with the baseline and decides on production.
API products, partner programs and open banking revenue
For heads of digital and business leaders who want APIs to earn revenue as well as meet open banking obligations.
We architected an API ecosystem that generates 8% of a bank's annual profit. We bring that experience in product design, partner onboarding, pricing and governance to your API strategy, whether the goal is open banking compliance that pays its way or a full partner ecosystem.
- API product strategy and portfolio design
- Partner programs and developer portals
- Open banking commercial models
- Usage analytics, billing and chargeback readiness
What you get
- API product catalog with target partners and use cases
- Commercial model for each product: pricing approach, plans and limits
- Partner onboarding process, from first contact to production
- Developer portal and sandbox, set up and documented
- Usage and revenue reporting design
How it typically runs
- Strategy 3 to 5 weeksPortfolio, partner segments and commercial models.
- Platform 6 to 10 weeksPortal, sandbox, plans and the onboarding workflow.
- Pilot partners First 3 monthsA small group of partners onboarded and measured.
- Scale QuarterlyNew products and partners on a regular cycle.
Products it draws on
- DevGateEarly accessAPI developer portal
- ConsentHubOpen banking platform
- OneSignCustomer identity and access
- DebitFlowDirect debit collections
Typical scenario
A bank meets its open banking obligations but earns nothing from its APIs. We select a few products with clear demand, such as recurring collections for corporate billers, define plans and onboarding, and launch them through a developer portal to a small group of pilot partners.
Test automation, simulation, performance and security testing
For delivery leaders and risk committees who need evidence that a regulated platform is ready for production.
A failure in production costs money, customers and the regulator's attention. We build the harnesses that produce release evidence: regression packs, simulations of the rails you cannot test against, and the security and performance proof your risk committee expects before go-live.
- Service virtualization: simulated core banking, payment schemes and government rails
- Recorded-traffic replay against real production messages
- Performance and load testing: TPS targets, soak tests, capacity evidence
- Security testing: API penetration testing, FAPI conformance, OWASP ASVS
What you get
- Test strategy and release criteria agreed with your risk function
- Automated regression and contract packs that run in your pipelines and stay with you
- Simulators and PDPL-safe synthetic data for core banking, payment and government interfaces
- Performance report against target TPS and latency
- Security findings with retest evidence, and a go-live evidence pack
How it typically runs
- Assess 1 to 2 weeksCurrent coverage, environments and release criteria.
- Build 4 to 8 weeksAutomation, simulators and test data in your pipelines.
- Prove Every releaseRegression, performance and security runs, with evidence.
- Hand over When readyYour team owns and runs the packs.
Products it draws on
- DataShiftReconciliation evidence for migrations
- ConsentHubFAPI 1.0 Advanced today, ready for 2.0
Typical scenario
A bank is replacing the integration layer behind its payment flows. Before cutover, recorded production messages are replayed through the new platform and each result is compared with the old one. Mismatches are fixed and replayed until every message matches, and the evidence goes to the change board.
How an engagement runs
Four stages. You can start at any of them, and each one ends with a decision you make before the next is committed.
Assess
Typically 2 to 4 weeks
We inventory what you run and map risk against your goals and regulatory deadlines.
- Integration inventory and dependency map
- Risk and rationalization analysis
- Costed, sequenced roadmap
Your decisionFund the roadmap, reshape it or stop. The outputs are yours either way.
Design
Typically 3 to 6 weeks
Target architecture, security model and operating model, decided and documented so they hold up under review.
- Architecture blueprint and interface specifications
- Security design aligned with SAMA, NCA ECC and PDPL controls
- Delivery plan with milestones and acceptance criteria
Your decisionApprove the design through your governance and agree the milestones.
Build
Every 2 to 4 weeks
Platform delivery in production increments, each tested against recorded production traffic.
- Working releases in production
- Test, performance and security evidence per release
- Runbooks and documentation kept current
Your decisionAccept each increment against the agreed criteria, then continue, reprioritize or pause.
Run
Ongoing, reviewed monthly
Managed 24/7 to bank-grade SLAs, or handed over to your own team.
- SLA and incident reporting
- Monthly service review
- Exit plan kept current
Your decisionKeep the service with us, bring it in-house or move it to another provider.
How we work with your teams and vendors
We work inside your governance and alongside the people and vendors you already have. Handover is planned from the first week.
Governance
- A named engagement lead and a steering committee with your sponsor
- Weekly status: progress, risks and decisions needed from you
- Work goes through your change, security and architecture approvals
- Risks, issues and decisions logged where your teams see them
Knowledge transfer
- Your engineers work alongside ours from the first sprint
- Documentation written as the work is done, in your repositories
- Recorded walkthroughs of each component for new joiners
- A named owner on your side for each platform before handover
Handover
- Handover pack: architecture, runbooks, test packs, access and certificates
- Source, configuration and pipelines held in your repositories throughout
- A shadow period in which your team operates and we stand by
- Sign-off against acceptance criteria agreed at the start
Your existing vendors
- A written split of responsibilities, agreed at mobilization
- Interface contracts and joint test plans at every shared boundary
- One joint plan and status report where several parties deliver
- Vendor evaluation and RFP support when you choose a new supplier
Questions executives ask before engaging
If your question is not here, send it to us and a senior practitioner will reply within one business day.
How are engagements priced?
Three models, matched to the work. Fixed-scope work, such as the estate assessment and the AI pilot, has a fixed price agreed before it starts. Platform delivery is priced by milestone, with payments tied to accepted increments. Managed services run on a monthly subscription with SLA credits. Product licensing, where relevant, is set out in the same written proposal.
Can you work alongside our existing vendors?
Yes, and you do not need to replace them to work with us. At mobilization we agree a written split of responsibilities with your integrators and vendors, then set interface contracts and joint tests at every shared boundary. Where one of our own products is an option in a vendor evaluation, we say so at the start.
Where is our data held?
In Saudi Arabia by default. Our platforms and AI solutions deploy inside your security boundary, on-premises or in your own cloud tenancy. Data handling is designed around PDPL and aligned with the NCA ECC and SAMA controls you report against. AI models run inside your environment, so operational data is not sent to shared external services.
How fast can you start?
The executive briefing can happen this week. Fixed-scope assessments start within two weeks of scoping, and AI pilots run for 30 days from kickoff to a go/no-go decision. Delivery and managed services typically start after two to four weeks of mobilization.
What support do we get after go-live?
Delivery engagements end with hypercare, typically four to six weeks, in which the delivery team stays on to fix defects and support your operators. After that you choose: your team runs the platform with the handover pack and runbooks, or we run it as a managed service with agreed SLAs and monthly reporting.
Who will we actually work with?
Senior practitioners who have owned these platforms inside a bank. The people in the briefing are the people on the engagement, and the work is not handed to a junior team after the sale.