AI that measurably takes work off your team
Secure, and fully local if you need it. For small and mid-sized businesses buried in email, proposals, documents and routine tasks: we find your most valuable AI opportunity, turn it into a working workflow within days, and then train your team. Delivered personally by Theo Gottwald – remotely or on site near Karlsruhe, Germany.
No documents needed · Non-binding proposal with price range · Your data stays with you
Five core services – book one or combine them
Start small, scale only when it pays off
Every step delivers a usable result on its own. You decide after each step whether and how to continue.
| Step | What you get | Price |
|---|---|---|
| Intro call | 20 minutes remote: goal, bottleneck and data situation, followed by a proposal with a price range | free |
| AI starter consultation | 60–90 minutes remote: which use cases pay off for you and where to begin | from EUR 290 |
| Prompt & workflow coaching | 2–4 hours working on your real tasks via screen sharing | from EUR 490 |
| Half-day workshop / team training | Your team learns safe AI use with its own examples and a prompt library | from EUR 900 |
| Workflow day | One concrete workflow in one day: email, proposals, documents, listings or recruiting | from EUR 1,500 |
| AI Opportunity Audit | 3–5 prioritised AI opportunities with business case, sprint plan and written decision brief | EUR 1,500–3,500 |
| AI Automation Sprint | 5 days of delivery: one production workflow including testing, handover and short training | EUR 4,500–9,500 |
Local-first path for sensitive data: LM Studio setup day from EUR 1,500 → MCP & WorkJinn integration from EUR 2,500 → governance workshop from EUR 3,500.
Custom development: time and materials, or a fixed price after a joint scoping session. All prices are guide prices; you receive a binding written quote before the project starts.
Business questions we turn into practical AI projects
You do not need to arrive with an AI specification. Bring us the work that consumes time; we turn it into a guided remote session, an on-site workshop or a focused implementation.
Amazon business support
Question: How can AI help run my Amazon business?
Tasks: Improve listings, analyze reviews, draft answers, research keywords, compare competitors and prepare stock or price lists.
Email and proposal assistant
Question: How can AI help us answer inquiries faster?
Tasks: Response drafts, reusable proposal sections, follow-up questions, summaries, follow-ups and quality checks.
Document review and summaries
Question: How can AI prepare contracts, PDFs, specifications or reports?
Tasks: Extract key points, flag risks, derive actions, compare documents and prepare a reliable handoff.
Recruiting preparation
Question: How can AI improve profiles, job ads and candidate communication?
Tasks: Sharpen role profiles, structure applications, prepare interviews, draft messages and assess matching criteria.
Real-estate workflows
Question: How do we save time on listings and prospect communication?
Tasks: Property descriptions, viewing preparation, prospect Q&A, owner reports and social content.
Support knowledge assistant
Question: How can our staff find reliable internal answers faster?
Tasks: FAQ structures, response templates, internal search, ticket triage, summaries and escalation rules.
Codex, local AI, LM Studio and MCP as bookable services
For customers who need more than a chatbot: practical AI workstations, developer workflows and local data processing, especially where sensitive data must not leave the organization.
Codex Workflow Setup
Setup and training for repository-aware development, planning, code changes, tests and reviews.
Local AI with LM Studio
Model selection, LM Studio setup, privacy boundaries and first workflows using your own files.
MCP & WorkJinn Integration
Connect AI to approved local files, Windows actions, data processing and controlled tool workflows.
Local-First AI Governance
Data classes, approvals, prompt rules, review controls, auditability and staff training.
Local AI and controlled tools when data cannot go online
Customer lists, contracts, technical documents, health information, internal calculations and NDA files may require a local operating model. We combine local AI, local files, controlled MCP tools and clear approval rules.
- 1Classify data: Decide what remains local and what may be processed externally after anonymization.
- 2Set up local AI: Install and test LM Studio or suitable local models.
- 3Control tool access: Give AI access only to approved tools, files and workspaces.
- 4Enable staff: Teach approved prompts, review rules and safe daily use.
Common use cases
Legal, healthcare, tax, engineering, HR, internal development and sensitive customer projects.
The goal is useful AI without losing control of data paths, files or tools.
Remote screen-sharing
We work together on real tasks in your shop, email, documents, spreadsheets, CRM, website, Windows tools or local AI environment.
Ideal for fast starts, coaching and focused two- to four-hour sessions.
On-site consulting
We map processes, train staff, build examples at the workplace and establish practical rules for management and teams.
Ideal for workshops, team training and local-first setups.
A fast path from problem to affordable project
- 1Free 20-minute intro call: Clarify your industry, goal, bottleneck and data constraints.
- 2AI audit: Prioritize the best three to five opportunities and estimate value.
- 3Automation sprint: Deliver one workflow in usable form.
- 4Training and support: Enable the team and plan the next improvements.
Best fit
Trades, agencies, legal and healthcare practices, property, recruiting, B2B services and software or Windows-heavy teams.
Especially valuable where inquiries, documents, email, reports or recurring office work create a daily backlog.
Theo Gottwald
Software developer and AI consultant based near Karlsruhe, Germany
- Developer of Smart Package Robot, the PBX compiler family and the SindByte MCP Server
- You talk directly to the person who delivers – no hand-offs, no call centre
- Local-first by conviction: our own software runs locally on Windows
What does the intro call cost?
Nothing. 20 minutes remote, no obligation. Afterwards you get a concrete proposal with a price range – or an honest answer that AI does not pay off for your case yet.
Does our data have to go to the cloud?
No. If you prefer, we work fully locally with LM Studio, local models and our MCP server. Where cloud AI makes sense, we agree together which content may only be processed in anonymised form.
Remote or on site?
Both. Remote via screen sharing for quick starts and coaching; on site near Karlsruhe for workshops, teams and sensitive processes. Other regions by arrangement.
What do you need from us?
One concrete bottleneck, a few sample emails or documents (anonymised is fine), a contact person and access to the tools involved.
When are we not the right partner?
For vague large-scale projects without data access, and for processes where nobody can check the result. We do not provide legal or tax advice.
What happens after the project?
You receive a handover, short training and documentation. If you like, we keep supporting the workflow and add further automations step by step.
Tell us which work AI should take off your plate
Two or three sentences are enough. You receive a time slot for the free intro call or a package recommendation with a price range.
Request a free intro call Get a package recommendationOr write directly to info@it-berater.org
