Our solutions
AI solutions for enterprise development and documentation processes
We integrate Gloster Digital Group’s two in-house AI products—the EstiMate AI project estimation system and the DocuLinq enterprise document management platform—into the client’s existing workflows. The systems run in European data centers and are tailored to the client’s infrastructure: on-premises, in a private cloud, or within the client’s Microsoft 365 environment.
As part of the development methodology
We have been developing enterprise software for more than two decades. We integrate AI tools into the design, estimation, documentation, and knowledge management phases just as we would any other engineering component: assigned to specific processes, with measurable outcome goals. A significant portion of Gloster’s clients operate in regulated industries—finance, insurance, utilities, manufacturing, government, and healthcare—so the baseline requirements for our solutions include GDPR compliance, auditability, and data retention at the client’s site.
We have taken our two in-house developed products—EstiMate AI and DocuLinq—from our own development and documentation practices and made them available at the platform level. Both continue to evolve in a live customer environment; updates are added to the roadmap based on usage data and customer feedback.
We have taken our two in-house developed products—EstiMate AI and DocuLinq—from our own development and documentation practices and made them available at the platform level. Both continue to evolve in a live customer environment; updates are added to the roadmap based on usage data and customer feedback.
Where does the value appear?
Based on our experience, the implementation of AI delivers measurable results in three key areas for Gloster’s client base. The specific savings depend on team size, processes, and the initial situation; we begin every implementation by assessing the baseline and measure the results against that baseline.
Time savings on routine tasks
Routine aspects of estimation, document search, and report analysis can be automated. This allows development teams and business units to devote more time to higher-value-added tasks—such as architectural decisions, client coordination, and risk analysis.
Data processing by the client
The client’s documents and processed texts remain within the customer’s infrastructure. Deployment is possible on-premises, in a private cloud, or within the customer’s Microsoft 365 / Azure tenant. Every query and model response is auditable, and access control aligns with existing permission management.
Data-Driven Estimation and Decision Preparation
Models that learn from actual data from past projects provide more reliable estimates than individual intuition. Once implemented, discrepancies—such as the difference between estimated and actual costs—can be tracked, allowing the planning process to be continuously calibrated.
Team of development engineers specializing in AI components
Gloster is primarily a software development organization: our accumulated knowledge and methodology stem from over 20 years of corporate delivery experience. We apply AI at those points in the process where a carefully calibrated model delivers measurably better results than the existing procedure. Our two dedicated products—EstiMate AI and DocuLinq—are the result of this approach: a specific, recurring business problem; a proprietary, working solution; and integration into the customer’s existing systems.
EstiMate AI
Project estimation based on the team's own data
INTRODUCTION
Software development project estimation revolves around three classic challenges: sprint planning often drags on for hours, story point estimates are frequently not objective, and post-project variances undermine client transparency. EstiMate AI provides an engineering solution to these pain points.
HOW IT WORKS
The system uses the team’s previous Jira tickets, sprint performance data, and the technical characteristics of tasks to estimate the expected effort for a new ticket. Alongside the proposal, it displays which previous tasks the ticket resembles and what factors have increased or decreased the estimate. Based on this information, the team makes an informed decision on whether to accept the proposal or override it.
BUSINESS IMPACT
- Shorter planning cycles. — The estimation portion of sprint planning is reduced to a matter of minutes; the meeting can then be devoted to discussing the details of the tasks.
- A more accurate quote. — Resource requirements for new projects can be estimated based on historical data, resulting in a more well-founded quote. This reduces the need for subsequent adjustments.
- Continuous calibration. — The system feeds data from every completed sprint back into the model; the accuracy of the estimates improves over time.
- Transparent decision-making. — The system displays the basis for each proposed estimate. The team can approve or adjust the value with full knowledge of the facts; the process is auditable.
TARGET AUDIENCE
- Development managers and tech leads — more accurate sprint commitments, fewer missed deadlines, and more measurable team performance.
- Project managers — data-driven scheduling and client communication; more well-founded proposal preparation.
- CTOs and delivery managers — comparable performance metrics across teams and projects; capacity planning can be calibrated based on this data.
INTEGRATION
Atlassian Jira and Jira Service Management. Installation is performed as an Atlassian Marketplace plugin. Initial setup involves processing the existing ticket backlog; Gloster provides engineering support for calibration.
DocuLinq
Enterprise document management with a search layer
INTRODUCTION
A significant portion of corporate knowledge is contained in unstructured documents—such as contracts, policies, reports, and correspondence. Locating, interpreting, and uncovering connections in these documents consumes a considerable amount of time on a daily basis; it also poses a risk prior to audits, legal reviews, or financial decisions. DocuLinq provides a solution to this problem.
HOW IT WORKS
DocuLinq is a layer built on top of existing document repositories—such as SharePoint, network drives, Microsoft 365, and access-controlled archives. The RAG (Retrieval-Augmented Generation) architecture enables the system not only to find documents containing a keyword, but also to answer the question in context by combining multiple sources. The source document, page number, and version appear next to each answer; the link leads directly back to the original file.
FOR WHOM AND FOR WHAT PURPOSE
CFO and Financial Management
Contract terms, payment deadlines, warranty obligations, and indemnification obligations can be queried in context—with a reference to the source contract. Comparing referenced historical financial reports and internal policies shortens decision-preparation time.
Compliance and Risk Management
In regulated industries—financial services, insurance, energy, healthcare, and pharmaceuticals—compliance checks can be performed using queries accompanied by an audit trail. DocuLinq records the source used, the query, and the user for every response; the audit trail can be exported.
Legal
When reviewing complex contract portfolios, the system highlights risk clauses, changes between versions, and inconsistencies between documents. The user continues to work with details traced back to the source rather than relying on the generated summary.
Business decision-maker
Organizational knowledge becomes available in a centralized, referenced format; the background material needed for decision-making can be gathered in minutes instead of hours.
Contract terms, payment deadlines, warranty obligations, and indemnification obligations can be queried in context—with a reference to the source contract. Comparing referenced historical financial reports and internal policies shortens decision-preparation time.
Compliance and Risk Management
In regulated industries—financial services, insurance, energy, healthcare, and pharmaceuticals—compliance checks can be performed using queries accompanied by an audit trail. DocuLinq records the source used, the query, and the user for every response; the audit trail can be exported.
Legal
When reviewing complex contract portfolios, the system highlights risk clauses, changes between versions, and inconsistencies between documents. The user continues to work with details traced back to the source rather than relying on the generated summary.
Business decision-maker
Organizational knowledge becomes available in a centralized, referenced format; the background material needed for decision-making can be gathered in minutes instead of hours.
SECURITY AND DATA MANAGEMENT
DocuLinq works in both on-premises and private cloud deployments. Document content and queries remain within the customer’s infrastructure. Processing takes place either via a model service running in a European region selected by the customer (typically the Azure OpenAI Service EU region, with contractual data processing guarantees) or via open-source models deployed in the customer’s environment; this is agreed upon with the customer as part of the solution architecture.
Access control uses existing Azure AD or other identity provider permissions, so there is no need for parallel permission management: only those who have permissions in the source system can access a document. Every query and response is logged with an audit trail.
Access control uses existing Azure AD or other identity provider permissions, so there is no need for parallel permission management: only those who have permissions in the source system can access a document. Every query and response is logged with an audit trail.
INTEGRATION
Microsoft SharePoint, Microsoft 365, and Azure AD; custom source systems can be integrated using custom connectors. The implementation begins with an assessment of the data sources and the coordination of the access model.
Data Management, Compliance, European Infrastructure
A significant portion of Gloster’s client base operates in regulated industries where controlled data management and regulatory compliance are essential components of AI implementation. Four key principles guide every implementation:
European data center, EU region
Customer data remains within European infrastructure—in Gloster Cloud’s own data centers, in a European-region Microsoft Azure tenant, or in the customer’s on-premises environment. The choice is made during the implementation planning phase, based on data classification and the compliance framework.
GDPR compliance from the planning stage
We design solutions based on the privacy-by-design principle: creating a data inventory, preparing legal basis documentation, and establishing access and deletion processes are all part of the implementation. The solution is delivered with a data processing agreement and technical documentation.
Auditability
Every model response, query, and model update is logged. The audit trail can be exported and integrated into the client’s SIEM or compliance system.
Industry-specific compliance frameworks
Regulatory requirements for the financial sector (MNB recommendations, DORA), healthcare (medical device validation frameworks), energy, and critical infrastructure (NIS2) are taken into account during the design phase. The exact scope depends on the client’s area of operation and the data processing risk level; this is documented separately for each implementation.
European data center, EU region
Customer data remains within European infrastructure—in Gloster Cloud’s own data centers, in a European-region Microsoft Azure tenant, or in the customer’s on-premises environment. The choice is made during the implementation planning phase, based on data classification and the compliance framework.
GDPR compliance from the planning stage
We design solutions based on the privacy-by-design principle: creating a data inventory, preparing legal basis documentation, and establishing access and deletion processes are all part of the implementation. The solution is delivered with a data processing agreement and technical documentation.
Auditability
Every model response, query, and model update is logged. The audit trail can be exported and integrated into the client’s SIEM or compliance system.
Industry-specific compliance frameworks
Regulatory requirements for the financial sector (MNB recommendations, DORA), healthcare (medical device validation frameworks), energy, and critical infrastructure (NIS2) are taken into account during the design phase. The exact scope depends on the client’s area of operation and the data processing risk level; this is documented separately for each implementation.
Where does the collaboration begin?
The implementation is preceded by a brief audit phase: we map out the processes, assess the current state, and work with the client to define measurable outcome goals. This is followed by a pilot deployment in a business unit or development team. Full-scale implementation takes place after the pilot results have been evaluated.
If the timing, scope, or business case for AI implementation is uncertain, it is advisable to start with this phase and only then select a specific product.
If the timing, scope, or business case for AI implementation is uncertain, it is advisable to start with this phase and only then select a specific product.
EstiMate AI
EstiMate AI is Gloster’s proprietary project estimation platform. It supports effort estimation for software development tasks based on the team’s historical performance data, resulting in a faster planning cycle, more objective estimates, and more predictable delivery. It operates within the Atlassian Jira environment, and the integration can be installed as a Marketplace plugin.
-p-500.png)
DocuLinq
DocuLinq is Gloster’s enterprise document management platform.It transforms unstructured document collections—such as contracts, policies, and reports—into a searchable, linked knowledge base with an audit trail. It is deployed on the customer’s infrastructure (on-premises or private cloud) or in a European Azure environment; the data remains under the customer’s control.
Get in touch with us!
How can Gloster contribute to your business success?
Contact us at info@gloster.hu or simply click on the link below!
Contact us at info@gloster.hu or simply click on the link below!
Gloster has established an abuse reporting system in accordance with Act XXV of 2023 on complaints, public interest disclosures, and rules relating to the reporting of abuse, which is available to all our employees, clients, contracted partners, and anyone else who comes into business contact with Gloster.
Reports can be made in the following ways: By email to visszaelesbejelento@gloster.hu
Reports can be made in the following ways: By email to visszaelesbejelento@gloster.hu
