01
Training
Your team practises using coding agents on everyday development tasks.
Together, we establish rules for using the tools and reviewing the code.
Augmentry / Engineering for business
We build software
and put AI
to work
We work with your business and engineering teams on existing applications and new projects, from technical decisions through to tested software.
Discuss your projectChristian Blank Software engineering & AI
We do not sell licences. We choose tools to suit your requirements. Our services
01 / Services
How we can help
You can bring us in for a specific task or a project. We agree on the scope and how we will work with your team before we begin.
02
Software development
We build new applications and extend existing systems, including AI integrations.
You receive tested code and the documentation your team needs to run and develop it further.
03
Audits and assessments
We examine software, architecture and technical risks against an agreed question.
Our report sets out the findings, our assessment and the limits of the investigation.
04
Automation
We automate workflows that currently involve copying data by hand or reviewing documents one at a time.
We implement error handling and document who is responsible for running the workflow.
05
Consulting
We help you make decisions about architecture, AI tools and building your own software.
We compare the options against your requirements and explain our recommendation.
06
Infrastructure
We plan and build the environment for your AI applications, in the cloud or on your own infrastructure.
With your IT team, we agree on data access, running costs and responsibility for operations.
02 / Topics
What is your team working through?
These pages explore the technical questions behind our services, with context and links to relevant support.
Agentic Engineering
How engineering teams can use coding agents consistently and establish shared practices for tasks, context, data and reviews.
AI Readiness and Governance
Assess existing capabilities and establish practical rules for data, tool access, approvals and technical responsibility.
AI Automation
Assess processes for useful automation based on data, interfaces, failure impact and operational effort.
AI-Assisted Software Delivery
Identify appropriate uses of AI in software development and adapt reviews, tests and technical responsibility accordingly.
AI Assurance and Assessments
We examine AI-related risks, systems and decisions from a technical perspective and document the results in writing.
Private AI Infrastructure
We determine whether private AI infrastructure is appropriate for your use case and which data, security and operational questions need to be resolved first.
Security Audits and Compliance
A traceable technical review of software, architecture, access controls and data flows, based on a clearly defined scope.
For your organisation
An established SME and a large enterprise work under different conditions. We take your existing teams, systems and requirements into account.
03 / Technical analysis
Do coding agents speed up development?
We explain the METR findings and why they cannot simply be applied to every team.
Coding agents · Productivity
Measure the whole job.
- Implement
- Review
- Operate
Rework feeds back
All three dimensions apply across the entire workflow.
- Time
- Active work, waiting and iterations
- Quality
- Defects, tests, review and maintainability
- System
- Flow, failures, recovery and team workload
Analysis of published research
Coding agents and productivity: what the METR studies actually show
The studies do not provide a universal productivity number for coding agents. They show how closely outcomes depend on tool maturity, task selection and measurement design.
Read our analysis ↗Prompt injection in a development agent: the VS Code attack chain as a system problem ↗
A safe reconstruction of the path documented by GitHub Security Lab and the technical controls required for agentic development environments.
GOV.UK Chat: what technical transparency looks like for a production AI system ↗
An analysis of GOV.UK Chat's public Algorithmic Transparency Record and what it contributes to governance, architecture and accountable operation.
04 / Knowledge & tools
Take a closer look
Our free checks and calculators help you assess a starting point. The guides explore individual questions in more detail.
EU AI Act Risk Checker
Screen concrete AI Act triggers, roles, requirements and dates with non-binding technical orientation.
AI Readiness Self-Assessment
9 questions show how well your organisation is prepared to use AI and where a closer review should begin.
Private AI Cost Calculator
Compare the monthly cost of a cloud API with self-hosting and calculate the break-even point.
Security Headers Check
Enter an address, check its HTTP security headers: a grade from A to F with concrete recommendations for every check.
AI Visibility Check
Check whether AI crawlers can access the public content on your website.
05 / About us
We are Christian Blank and Simon Daum
Christian has worked in software engineering for more than twelve years, including medical technology, pharmaceuticals and industry.
Simon develops web and mobile applications and automates business workflows.
06 / Working together
From the first question to handover
Not every engagement needs all four stages. We agree on where to begin and what you should receive at the end.
- 01 / Assess
Understand the task
We examine your systems and talk to the people involved, then record the scope and any questions that still need answers.
- 02 / Equip
Prepare the work
We choose tools and architecture to meet your requirements, and agree on working practices and decision-making with your team.
- 03 / Embed
Build and review
We deliver the agreed work in stages so your team can see and review the results with us.
- 04 / Sustain
Hand over the system
We document the solution and walk your team through it, recording responsibilities and outstanding work at handover.
Contact
Tell us about the task and where you need support. We will discuss whether and how we can help.
What are you working on?