We are looking for a CTO to lead the technology function of a trading business, with responsibility for architecture, software development, infrastructure, data, AI adoption and growing the engineering team.

At the core of the role are systems with high event throughput, real-time data processing and strict requirements for latency, availability and correctness of operations. We need a leader who can tie technical decisions to business outcomes: keep the trading platform running reliably, accelerate product development, make data trustworthy and adopt AI with measurable impact.

The CTO must be comfortable operating at two levels: setting the technology strategy while maintaining a deep understanding of architecture, performance, incident root causes and the quality of engineering decisions.

Core mission of the role

Build and evolve a technology platform that sustains growing trading activity, preserves the correctness of operations under peak load and during failures, provides the business with reliable data and enables changes to be shipped quickly with controlled risk.

The CTO’s work is expected to deliver:

A reliable and well-controlled trading platform.
Predictable performance as load grows.
A transparent architecture with clear boundaries of ownership.
A strong engineering team and sustainable development processes.
Reliable analytics for operational and strategic decisions.
Working AI use cases with proven economics.
Controlled infrastructure, development and compute costs.
A technology strategy aligned with the business plan.

Key responsibilities

1. Technology strategy and platform development

Develop a 12–18 month technology strategy and a detailed work plan for the next six months.
Audit the architecture, infrastructure, development processes, security, data and team capabilities.
Identify constraints that block scaling, slow down product development or create financial and operational risks.
Define the target architecture and the migration path towards it while ensuring business continuity.
Prioritise technical initiatives by their impact on reliability, delivery speed, data quality, operating costs and commercial objectives.
Manage technical debt: document its consequences, estimate the cost of remediation and include the necessary changes in the overall plan.
Make build, open-source and buy decisions for external services.
Assess the total cost of ownership of technologies, vendor lock-in risks and the complexity of long-term maintenance.
Ensure architectural decisions match the actual requirements of the business, the load and the team.

2. Trading systems architecture

Own the technical architecture for market data ingestion, trading event processing, order management and integrations with external venues.
Ensure correct handling of the order lifecycle: creation, submission, acknowledgement, partial fill, cancellation, rejection and state recovery after a failure.
Build in safeguards against reprocessing, duplicates, out-of-order events and ambiguous responses from external systems.
Set up reconciliation of orders, trades, positions and cash balances between internal systems and external sources within the scope used by the business.
Ensure controlled behaviour on connection loss, delayed market data, API rate limits and venue unavailability.
Together with the owners of the trading and risk functions, implement technical limit checks, access restrictions and kill-switch mechanisms.
Ensure audit logging of financially significant actions and the ability to reconstruct the sequence of events.
Create the technical conditions for reproducible research, testing and controlled rollout of trading algorithms, where the business uses them.
Separate the research environment, testing and production; define the rules for admitting changes into the trading platform.

3. High load and low latency

Define throughput, latency, data volume and acceptable error requirements for each critical scenario.
Measure performance across the entire processing chain, including network interactions, queues, databases and external integrations.
Manage tail latency: p95, p99 and, where necessary, p99.9.
Identify and eliminate bottlenecks in CPU, memory, garbage collection, disk I/O, networking and resource contention.
Design the system to handle sudden event bursts, queue build-up and recovery from overload.
Implement load shedding, backpressure, prioritisation and graceful degradation.
Organise reproducible load tests with profiles that closely reflect real-world operation.
Plan capacity based on forecast growth and the required headroom.
Choose technologies and approaches based on measurements, correctness requirements and maintenance costs.

4. Reliability and operations

Define the list of critical services, their owners and target service quality levels.
Implement SLI/SLO: measurable indicators of availability, latency, errors, data freshness and processing correctness.
Build observability based on metrics, logs and tracing that makes it possible to localise problems quickly.
Configure alerts that reflect the impact on users and business processes.
Organise on-call rotations, escalations, runbooks and allocation of responsibilities during incidents.
Run post-incident reviews of major incidents with root cause analysis and follow-up on corrective actions.
Define service recovery requirements and acceptable data loss for different classes of systems.
Organise backups and regularly test restores with verification of data integrity.
Test scenarios of infrastructure failure, loss of a dependency, data corruption and recovery of trading integrations.
Ensure safe rollout of changes, the ability to roll back and controlled disabling of problematic functionality.

5. Artificial intelligence and machine learning

Identify priority AI use cases in trading research, analytics, information processing, operational processes and software development.
For each use case, define the expected impact, quality requirements, implementation cost and rules for evaluating results.
Choose between external models, self-hosting, fine-tuning and traditional algorithms based on the task and its economics.
Organise training and evaluation data management: provenance, quality, access, versioning and reproducibility.
Ensure correct separation of training and validation data, including time-based splits for trading tasks and prevention of look-ahead leakage.
Establish a process for validating models before deployment and monitoring quality after launch.
Track quality degradation, input data drift, compute costs and latency.
Define the permitted actions of AI, the need for human review and the conditions for automatic shutdown.
Ensure accounting for the costs of models, infrastructure, human review and error correction.
Expand the use of AI within the engineering team, assessing its impact on development speed, quality and cost.

6. Data and analytics

Design the architecture for data collection, processing, storage and delivery.
Define requirements for stream and batch processing, change history and recalculation.
Ensure data consistency across trading, financial, operational and analytical systems.
Agree on definitions of key metrics with the business and assign their owners.
Set up checks for data completeness, freshness, uniqueness and consistency.
Ensure metric lineage: from a report down to the source events and calculation rules.
Build an analytics system accessible to management covering trading operations, fees, execution quality, risks, costs and platform performance, in line with the business model in use.
Provide controls over corrections, recalculations and methodology changes.
Enforce data access controls and ensure auditing of access to sensitive information.
Reduce the dependence of management decisions on manual exports and inconsistent spreadsheets.

7. Information security

Define the access model for infrastructure, data, trading integrations and administrative functions.
Ensure least-privilege access, strengthened protection of privileged accounts and controlled access provisioning.
Organise management of secrets, integration keys and certificates, including their rotation.
Embed security checks into development and release processes.
Ensure control over dependencies, vulnerabilities and remediation timelines for significant risks.
Organise auditing of critical actions and protection of logs against unauthorised modification.
Define the security incident response procedure and the escalation path to the responsible executives.
Together with subject-matter specialists, take into account data processing and storage requirements applicable to the company’s markets and jurisdictions.
Maintain a technology risk register with owners, deadlines and agreed mitigation measures.

8. Engineering management

Build a transparent process from task definition to release and verification of results in production.
Ensure realistic planning that accounts for available team capacity, operational load and technical risks.
Set definition-of-done criteria for tasks, architectural changes and releases.
Develop CI/CD, automated checks and testing of critical scenarios.
Organise code review, architecture discussions and documentation of significant decisions.
Reduce change lead time without compromising reliability and correctness.
Manage dependencies between teams, external vendors and business units.
Ensure delivery against commitments is measurable and the reasons for deviations are transparent.
Maintain up-to-date documentation sufficient for operations and handover of ownership.

9. Team and organisational structure

Design the structure of the engineering organisation based on current objectives and growth plans.
Define the areas of responsibility of managers, tech leads and teams.
Draw up a hiring and capability development plan across software development, infrastructure, data, AI and security.
Take part in hiring key specialists and assessing engineering managers.
Develop managers and build a succession pipeline.
Provide regular feedback and performance reviews.
Reduce the dependence of critical systems on any single employee.
Create the conditions for knowledge sharing, high-quality onboarding and sustainable on-call teams.
Evaluate employees based on results, quality of decisions and collaboration.

10. Budget and business partnership

Manage the technology budget within the agreed delegated authority.
Plan spending on infrastructure, licences, market data, external services and AI compute.
Calculate the unit cost of useful workload for the main technology domains.
Justify investments through expected impact, risk reduction or removal of growth constraints.
Monitor fulfilment of obligations by technology vendors and contractors.
Regularly report to management on platform status, timelines, risks, budget and decisions required.
Align priorities with the product, trading, finance and risk functions.
Flag constraints in a timely manner and propose courses of action with an assessment of their consequences.

Authority and scope of responsibility

The CTO is accountable for the technology strategy, architectural decisions, organisation of engineering work, operational quality and delivery of the agreed plan.

Within the approved budget and governance structure, the CTO:

Sets technology standards and architectural principles.
Prepares proposals on team structure and hiring.
Allocates engineering resources according to agreed priorities.
Approves the technical readiness of critical changes.
Initiates a review of timelines and scope when significant risks are identified.
May halt the release or operation of a technically hazardous change in accordance with the established procedure.

Trading strategies, acceptable financial risk and capital management rules are agreed with the owners of the respective functions. The CTO ensures the technically correct enforcement of these rules, the reliability of data and the resilience of the platform.

Candidate requirements

Leadership experience

Experience leading an engineering organisation as CTO, VP Engineering, Head of Engineering or in a comparable role.
Experience managing tech leads and multiple development streams.
Hands-on experience in hiring, team building, goal setting and performance evaluation.
Ability to build relationships with the business, discuss priorities and advocate for necessary technical investments.
Experience managing budgets and external technology vendors.
Ability to make decisions with incomplete information and limited resources.

Technical expertise

Deep understanding of distributed systems, networking, concurrency and consistency models.
Experience designing and operating high-load systems with measurable latency and reliability requirements.
Hands-on experience with event streams, queues, databases, caching and real-time processing.
Ability to analyse performance and identify the causes of degradation at the application and infrastructure levels.
Understanding of idempotency, event ordering, redelivery and recovery from partial failures.
Experience in setting up observability, CI/CD, incident management and disaster recovery.
Understanding of security engineering fundamentals.

Trading and data

Experience in trading, fintech, payment systems or another business with critical financial operations.
Understanding of the order lifecycle, trade execution, positions, reconciliations and the constraints of external integrations.
Experience building analytics systems and ensuring data quality.
Ability to investigate discrepancies between business metrics and technical sources.

AI and machine learning

Hands-on experience implementing AI or ML in business workflows, or leading such projects.
Understanding of quality evaluation, model limitations, operating costs and risk control.
Ability to determine where AI delivers measurable impact.
Understanding of the difference between a successful experiment and a system ready for continuous operation.

Nice to have

Experience in algorithmic trading, exchange integrations and market data processing.
Experience with low-latency systems and tail latency optimisation.
Deep expertise in Go, Rust, C++, Java or another language used in high-load systems.
Experience building data platforms, MLOps and model infrastructure.
Experience taking a critical platform from instability or rapid growth to stable operation.

Six-month work plan

Month one — audit and baseline

We expect:

A map of systems, integrations, data flows and owners.
An assessment of the team and capability gaps.
A register of the main technology risks.
Baseline metrics for availability, latency, load, data quality, delivery speed and costs.
A list of critical trading scenarios.
Proposals for priority changes.
An agreed KPI measurement methodology.

Outcome of the month: management understands the state of the technology, the main constraints and the course of action.

Month two — strategy and basic control

We expect:

An approved roadmap with budget, dependencies and owners.
Agreed SLOs, recovery requirements and incident response rules.
Closure of the most critical gaps in access control, backups and observability.
A prioritised hiring plan.
Approved definitions of the main business metrics.
Selection of AI use cases and criteria for assessing their value.

Outcome of the month: key risks, workstreams and metrics are under regular control.

Month three — first proven improvements

We expect:

First changes that remove significant performance or reliability constraints.
A working release management and post-incident review process.
Load testing of priority scenarios.
The first restore test of critical systems.
Automated quality checks for key data.
Validation of the first AI use case on representative tasks.

Outcome of the month: the impact of technical changes is confirmed by measurements.

Month four — developing the core domains

We expect:

Delivery of priority architectural changes.
Expanded automated reconciliation and control of trading operations.
A working system of core management metrics.
Elimination or reduction of critical key-person dependencies.
Launch of a suitable AI use case in limited production, provided the quality criteria are met.
First confirmed results of unit cost optimisation.

Outcome of the month: the main initiatives are delivering impact in production.

Month five — resilience testing

We expect:

Testing of the platform under forecast peak load with the agreed headroom.
Failure, recovery and kill-switch drills.
Removal of the identified constraints.
Confirmation of analytics quality over a regular cycle.
Accumulation of data on the impact of the deployed AI.
A practical test of backup ownership for critical services.

Outcome of the month: the resilience of the solutions is confirmed by testing and production operation.

Month six — evaluation of results and next phase

We expect:

A final KPI assessment backed by supporting data.
A comparison of results against the baseline.
A report on budget execution and changes in the cost of core operations.
An assessment of remaining risks and technical debt.
A team and platform development plan for the next period.
Agreed priorities for further scaling.

Outcome of the month: the business receives a transparent assessment of the impact achieved and a well-founded plan for the next phase.

CTO KPIs for the first six months

The numerical values below are a starting draft of the targets. During the first 30 days, the baseline metrics, measurement boundaries, critical systems, available resources and final acceptance criteria are established.

If the company’s existing requirements are stricter than those proposed, the stricter requirements apply.

Area	Weight	Target outcome by the end of month six
Platform reliability	15%	SLOs, owners, monitoring and response procedures are defined for all critical scenarios. The agreed availability has been achieved over the last eight weeks; the initial benchmark is at least 99.95% of valid operations processed technically correctly. A review has been conducted and corrective actions assigned for every major incident.
Performance and scalability	15%	For at least two selected critical flows, p99 is reduced by 25% relative to the baseline, or compliance with a pre-agreed sufficient absolute latency budget is confirmed. Testing confirms operation at 150% of forecast peak load without breaching correctness, latency or error requirements.
Trading platform correctness	15%	100% of agreed critical order processing paths have passed acceptance for state control, duplicate protection, applicable limits and auditing. Automated reconciliation of critical trading data is in place. There are no unexplained discrepancies outstanding beyond the agreed SLA. At least two kill-switch and integration recovery drills have been conducted.
Security and recovery	10%	Access requirements, RTO and RPO are defined for 100% of critical systems. Restores have been successfully tested with verification of integrity and business state. There are no overdue critical risks without formal acceptance by an authorised owner. Privileged access and critical secrets have been brought in line with the agreed management model.
Delivery predictability	10%	At least 85% of weighted priority commitments have been delivered on the agreed timelines and accepted against the definition of done. Median lead time for comparable tasks from ready-for-development to production has been reduced by at least 20%. The share of changes causing an incident, rollback or hotfix stays within the agreed limit and is not compromised for the sake of speed.
AI impact	10%	At least two priority use cases have been validated and at least one has been put into production after quality and economic review. A reduction in effort of at least 15%, or another pre-agreed measurable impact, has been confirmed while maintaining quality requirements. An owner, monitoring rules and shutdown conditions have been assigned.
Data and analytics	10%	Definitions, sources and owners are approved for 100% of agreed key metrics. Quality and freshness checks are in place for critical data. Over the last eight weeks, at least 99% of scheduled checks have met the agreed criteria. Core management metrics are available to management without regular manual compilation.
Team and ownership	5%	Every critical service has a primary owner and a trained backup. At least 80% of approved priority vacancies have been filled, given the allocated budgets and agreed hiring terms. Areas of responsibility and goals are defined for key employees; a feedback cycle has been completed.
Infrastructure and compute economics	5%	The unit cost of useful workload has been calculated for the core domains. In the selected areas it has been reduced by at least 10%, or a pre-approved target level has been met. Reliability and quality requirements have been maintained. Spending deviation from the agreed forecast over the last two months does not exceed 10% with load within the planned range.
Strategy and technology risk mitigation	5%	The audit and risk register are presented by day 30; an agreed roadmap with budget and dependencies by day 60. By the end of the six-month period, at least 90% of weighted priority measures to mitigate the most significant technology risks have been completed. The outcome is confirmed by working changes and acceptance.
Total	100%

KPI measurement and evaluation rules

Baseline and data sources

For each KPI, the following are established by the end of the first month:

Starting value.
Target value and deadline.
Measurement scope and list of systems.
Data source.
Person responsible for confirming the result.
Required resources and external dependencies.
Rules for partial achievement.

If measurements are not available, the CTO arranges for their collection and agrees on a sufficient observation period. Relative improvement is calculated only from a confirmed baseline value.

Availability

Availability is calculated as:

Valid operations processed technically correctly / all valid operations × 100%.

Internal errors, timeouts and overload rejections count as failed processing. A correct rejection due to an established risk limit counts as a technically correct result.

Metrics for the in-house platform and for the end-to-end scenario including external venues are reported separately. Technically correct processing of an order does not imply that the trade will necessarily be executed by the market.

Performance

Comparisons are made between the same classes of operations, comparable data volumes and pre-defined load profiles.

The report includes latency, useful throughput, errors, retries, queue depth and resource utilisation. Results must not be improved by excluding slow operations, dropping requests or reducing the useful workload.

Latency, capacity and cost targets are agreed together, specifying the budget and the test configuration.

Plan delivery

Delivery is calculated as:

Sum of weights of accepted commitments / sum of weights of all commitments for the period × 100%.

Weights are set before work begins, based on business priority. Work is considered complete when it meets the acceptance criteria.

Changes to timelines and scope are recorded promptly, with the reasons stated. The original plan is retained for variance analysis.

AI impact

Impact is assessed on comparable tasks against the original process or a control group.

The costs of models, infrastructure, implementation, human review and error correction are taken into account. Time savings count only if the agreed quality level is maintained.

For use cases that affect trading decisions, risk criteria and production admission criteria are agreed separately.

Data quality

The check success rate is calculated as:

Checks meeting the established criteria / all scheduled checks × 100%.

Skipped or non-functioning checks count as failed. Critical discrepancies in trading operations and balances are assessed separately: they cannot be masked by a good average across the rest of the data.

External dependencies

Changes in trading volume, venue unavailability, delays in budget decisions and hiring freezes are reflected in reporting.

Revising targets requires approval, and the history of changes is retained. The CTO is responsible for the response measures to external failures provided for in the architecture and for timely communication of risks.

Final assessment

The overall result is calculated as the sum of weighted scores across the ten areas. For each area, thresholds for full, partial and non-achievement are set in advance.

Mandatory requirements for the correctness of trading operations, access control and recovery are assessed separately. Failure to meet them cannot be offset by cost savings or faster delivery.

The link between the KPI assessment and variable compensation, if applicable, is set out in a separate agreement.

Reporting to management

Weekly: status of priority workstreams, significant risks, incidents and decisions requiring business input.
Monthly: metrics on reliability, performance, delivery, data, AI, team and budget.
For major incidents: prompt notification, impact assessment, root cause analysis and a plan to prevent recurrence.
At the end of the six-month period: assessment of KPIs, impact achieved, remaining constraints and the next phase of development.

Terms

Compensation: €15,000 per month.
Role level: CTO / Chief Technology Officer.
Scope of responsibility: technology, architecture, software development, infrastructure, data, AI and the engineering team.
Employment type, tax treatment of the amount, location, working arrangements and any variable component are agreed separately.
Authority, budget and available resources are agreed at the same time as the targets are approved.

What we expect from candidates when applying

A CV and several specific examples of your work:

Which systems and teams you have led.
What load, latency and availability requirements you have worked with.
What architectural or organisational changes you have implemented.
How the metrics changed before and after those changes.
What experience you have in trading, financial operations, AI and analytics.
How you handled major incidents and what systemic changes you made afterwards.

Confidential figures may be provided as ranges or in anonymised form.