How we work

Enterprise AI projects in three phases, from fixed-price analysis to handover

Every theBlue.ai project runs in three phases: a fixed-price process analysis, a milestone-based implementation on your real data and systems, and a handover of code and documentation. You can start at any phase.

Start with a process analysis See three projects

Some of the organisations we have built for

How we work

Three phases. Clear scope at every step.

We don’t start with a big contract. We start with a focused analysis and you see exactly what’s possible and what it takes before committing to anything.

Phase 1 · Your starting point

Process Analysis & Architecture Proposal

We work with your team, interviewing stakeholders and mapping workflows end to end. That includes the systems the process runs through, from SAP and other ERP systems to Microsoft 365 and your ticketing tools, and the data flows and security requirements around them.

On-site or remote sessions with your operations, IT, and domain teams
Whether your data carries the case, and what today’s models can do with your material
Architecture proposal naming the models, the retrieval and the agents, designed for your security standards and internal review
Clear scope, timeline, and cost estimate for implementation

What you get

A complete proposal covering the AI architecture, the integration plan, timeline and cost. It is ready to present to internal stakeholders, IT, and security. You decide whether to proceed.

Fixed priceStandalone engagement No commitmentTo proceed further

→ Your internal review: stakeholder alignment, security approval, budget decision

Phase 2 · Build & integrate

Development & System Integration

Based on the approved architecture, we build the AI system and integrate it into what you already run: your ERP, ticketing, email and document systems. Development happens iteratively, you see working components early.

Integration respecting your security and compliance requirements
Regular checkpoints so your team can review and give feedback
Edge cases refined together with your domain experts
Scope managed through milestones, and if requirements shift we re-scope transparently

What you get

A working system integrated into your environment, tested against your real process requirements and ready for production validation.

Milestone-basedPricing IterativeDevelopment

→ You can start at this phase: with a scope and an architecture of your own, or a prototype that has to hold up on real data. We review what you have, then build against your plan.

Phase 3 · Go live & handover

Production Rollout & Ongoing Support

The system goes live with real data. We monitor closely, refine based on real-world performance, and hand over with full documentation and training.

Controlled go-live with real workflows
Monitoring and rapid iteration during the initial period
Full documentation, training, and knowledge transfer

What you get

A production system your team uses daily, with full ownership, documentation, and ongoing support available.

Full ownershipCode and documentation Ongoing supportOptional

Phase 1 is a standalone engagement. You receive the full analysis and architecture proposal regardless of whether you proceed with implementation. Most clients use the Phase 1 deliverable to get internal alignment and security approval before moving forward.

See the individual services

Proof

Three projects that went into production

Three industries, three different starting points: a process still done by hand, an existing LLM system taken the last stretch to production, and a team losing its best hours to routine document work.

Security questionnaires answered in one week instead of one month

MedTech · apoQlar

Security questionnaires answered in one week instead of one month

Every new hospital client required a completed security questionnaire: eight to ten people, about a month, answers pulled from policies across departments. A GenAI assistant now drafts the responses and a small team verifies them.

Manufacturing · Radaway

Order data read straight out of customer email

Orders arrive as free-form email in whatever wording the customer chose. We took an existing LLM system the last stretch to production: semantic product matching, attachment processing and 90 percent less manual work.

Behind one orchestrator

Public affairs · RPP Group

Five specialised agents that work like an extra team member

Senior consultants spent hours on document analysis, data structuring and drafting, work that crowded out the strategic side. ChatRPP now carries it, in the firm’s own tone and inside their own infrastructure.

Read the case study

See more case studies

Testimonials

What clients say when the work is done

Healthcare · Hospital operations

An AI system that anonymises patient data on the hospital’s own servers

100%Manual anonymisation effort eliminated

On-premRuns on hospital servers, no cloud

Read the case study

The growing need for anonymization had become a time-consuming manual process. Together with theBlue.ai, we were able to transform this challenge into an efficient, AI-powered process, significantly reducing manual effort and relieving the burden on our staff.

Pietro Lucillo, MA IT Project Manager, Tirol Kliniken

MedTech · Compliance

Security questionnaires answered by retrieval over your own documents

–75%Completion time, a month to under a week

$90KEstimated annual saving

Read the case study

Managing the completion of security questionnaires is no longer a logistical nightmare. The new system is easy to manage and ensures our responses are accurate and comprehensive.

Maciej Antoszczuk Tech Product Owner, apoQlar

Government & public affairs

An AI system that reads millions of political documents

MillionsPolitical documents processed

PoC → ProdProof of concept to live product

Read the case study

Our close collaboration with theBlue.ai was key to unlocking the potential of AI for our needs. From the initial proof of concept through to the product launch, their structured and agile approach kept things efficient and on schedule. Early customer feedback has been incredibly positive.

Marc-Angelo Bisotti CEO, Policy-Insider.AI

Professional services · Copilot enablement

A Microsoft Copilot workshop for daily workflows

CopilotWord, Excel, Teams and Outlook

Multi-deptDepartments trained together

Read the case study

The training was conducted in an engaging way and tailored to our needs, resulting in very high ratings for both the content and the delivery. We recommend Apollogic and theBlue.ai as reliable training partners.

Marek Strzelczyk CIO, GS1 Poland

Your side of it

What we need from you, and from whom

The first question a project lead gets asked internally is whose time this will take. So here it is, before you have to ask.

People who run the process

One or two of them in the workshops, including whoever knows the exceptions nobody wrote down.

An operational lead

Someone who can decide what is in scope, and who stays reachable between sessions.

A contact from IT or security

For one session on the systems involved, from SAP to Microsoft 365, their interfaces and what your rules allow.

Real documents

From the process, anonymised if you like, with the mess still in them.

No surprises

The risks you carry, and where each one is handled

Scope, security and daily use decide how a first AI project ends. Here is where each one sits in the three phases.

Security has to approve it

Your IT and security people join in the first week of the analysis, so the architecture is drawn around your rules while it is still on paper. On-premise, cloud and hybrid are all routes we build on, and the proposal is written to go straight into your internal review.

Handled in phase 1

Requirements will change

They always do. The work runs in milestones, so a change is re-scoped in the open and its effect on time and cost is agreed in writing before the next milestone starts. Your budget moves when you sign.

Handled in phase 2

The team has to use it every day

The people who run the process today are part of the analysis, so the system fits the way the work is actually done. Go-live is controlled and runs on real workflows, the first exceptions are handled while we are still on the project, and training and documentation come with the handover.

Handled in phase 1 and 3

Questions about the process

What people ask before they sign

Can we skip the analysis if we already have a plan?

Yes, you can skip the analysis if you arrive with a defined scope, a use case and an architecture: theBlue.ai then starts at implementation. We review what you have first and tell you where we would do it differently, but the plan stays yours. Skipping the analysis is normal when the thinking has already been done, either by your own team or by someone else.

We already built something and it is not working properly. Can you fix it?

A prototype that fails on real data can usually be fixed without starting again. Prototypes that work in controlled conditions and fail in production are a common pattern, and the cause is normally in a few identifiable components. We review what exists, say what is salvageable, and rebuild only the parts that need it. That is how we took an existing LLM system at Radaway into production in three weeks.

Can we pause between the phases?

Yes, you can pause between the phases, and plenty of clients do. The analysis is a standalone engagement, so a pause after it is the normal case. Budget cycles and internal approvals take the time they take, and the analysis does not expire.

We already have several AI initiatives running. Where do we start?

When several AI initiatives are already running, the first piece of work is a roadmap rather than another build. Where efforts sit in several business units, each with its own tooling, it establishes what runs today, what overlaps, what earns its place, and where the next investment goes. The three phases then run on whichever process comes out of that first.

How much of our team’s time does the analysis take?

The workshop days are the bulk of it, and they are spread out so nobody has to abandon their job for a week. Between sessions we need one reachable contact for questions rather than a standing commitment. The number of days depends on how many departments and systems the process crosses, and we agree it before we start.

Do the workshops have to be on site?

They work remotely and we run them that way when travel does not make sense. On site is usually better for the first block, because half of what we learn comes from watching someone do the work. After that, remote is fine.

What happens to the analysis if we do not build with you?

The analysis is yours to keep whether or not you build with theBlue.ai. The process map, the architecture direction and the cost estimate are written to be used, including by someone else. Some clients take it to their internal team, some put it out to tender. We would rather that than hold a document hostage.

What does the proposal look like as a document?

A written document, structured so it survives being forwarded: the process as it runs today, the architecture we propose, the systems it would touch, a timeline with milestones, and a cost estimate. It is written for people who were not in the workshops, because those are usually the people who decide.

Which of our systems does the analysis look at?

The analysis looks at every system the process touches, which is usually an ERP such as SAP, Microsoft 365 and email, a ticketing tool, and wherever the documents are stored. theBlue.ai maps how data moves between them and what your security rules allow, and the architecture proposal is written to plug into those systems rather than replace them.

Can we start smaller than a process analysis?

Yes, a hands-on workshop is a smaller starting point than a process analysis. A session on ChatGPT, Copilot, Gemini and Claude for one department is a smaller commitment, and no project has to follow it. Some companies start there, find their own candidates for automation, and come back with a process in mind.

Tell us which process is costing you the most

Describe the process and we’ll come back within one business day with an initial assessment and a proposal for a 30-minute scoping call.

Phase 1 is a standalone engagement, with no commitment to proceed further.
Senior engineers from day one. No account managers, no handoffs.
We integrate, never replace. Your systems stay exactly as they are.

Do not fill this out:

NAME*

COMPANY*

EMAIL*

PHONE (OPTIONAL)

WHERE DOES THE MANUAL WORK SIT?*

Not sure which process to start with? Describe where the time goes and we work that out together.

Data Controller Information

The controller of your personal data is theBlue.ai GmbH, headquartered in Hamburg, Germany. By submitting this form, you consent to the processing of your personal data for the purpose of responding to your inquiry. You may withdraw your consent at any time, without affecting the lawfulness of processing based on consent before its withdrawal. Based on our legitimate interest, we may also send you information about our services and solutions, but only if it relates to the topic of your message. If you prefer not to receive such communications, you have the right to object at any time. For more details on how we handle your personal data and your rights, please refer to our Information Clause and Privacy Policy.

* Required fields.

We’ll come back within one business day.