Mathias (Dr.) Wilhelm-Development of an AI-driven social media automation for identifying topics, generating text, and publishing content
Check rate
Experience
Implementation of an on-premise OCR solution with information extraction
Mindhopper GmbH
- Insurance service provider*
Challenge: Business-critical documents were processed through external OCR providers, with ongoing costs, dependency, and data privacy risks for sensitive insurance data.
Implementation:
- Architecture and production implementation of an on-premise OCR solution with full data ownership
- Methods for recognizing document structures as the basis for automated further processing
- ML-, NLP-, and LLM/VLM-based information extraction, especially from invoices and quotations
Success: Replaced external providers: full data ownership, GDPR-compliant processing, and 75% lower recurring OCR costs per year
Used technologies: Python, Docker, Microservices, FastAPI, PyTorch, Torchvision, MongoDB, MySQL
Modernization of the data platform and BI reporting
DearEmployee GmbH
- Service provider for corporate health tracking*
Challenge: BI reporting was based on a manual, undocumented ETL process with inconsistent data and 13-second dashboard loading times; changes were time-consuming and error-prone.
Implementation:
- Analysis of the data landscape and Power BI data models
- Setup of a privacy-compliant data warehouse with ETL pipelines (SQLMesh); privacy-compliant data models and reporting logic for sensitive corporate health data (aggregation, role-based access)
- ML-based enrichment of user data, including derivation of demographic features
- Development of a new dashboard
- Design of an MCP server to provide DWH and BI data for text-to-SQL and future agentic AI applications
Success: Automated, documented data pipeline as a single source of truth, dashboard loading time reduced from 13 to 3 seconds, solid foundation for agentic AI applications
Used technologies: Scikit-learn, Python, PostgreSQL, ClickHouse, SQLMesh, FastAPI, React, Power BI
Automation of social media marketing
nuco.cloud
- Cloud computing provider*
Challenge: Manual content creation tied up significant resources – with irregular posting and limited audience-specific messaging.
Implementation:
- Analysis of marketing processes and identification of automation potential for each channel
- Agentic AI system for identifying company USPs, target audience analysis, and trend research, taking GDPR, transparency requirements, and human approval processes into account
- End-to-end implementation of AI-supported generation, planning, and cross-channel publishing
- RAG- and MCP-based knowledge layer for internal company information and external news sources
Success: Tripled posting frequency with much lower manual effort and more consistent, target-group-focused communication
Used technologies: LangChain/LangGraph, PGVector, PostgreSQL, Python, FastAPI, GitHub Actions, Docker
Development of an AI-driven social media automation for identifying topics, generating text, and publishing content
- Optimization and automation of marketing activities across social media channels
- Capture of requirements and the processes to be optimized
- Creation of an automation to identify company USPs for different marketing target groups
- Development of an algorithm to search for and identify relevant news
- Implementation of an algorithm to create and send social media posts
- Design and implementation of the complete automation in Python
- Creation of a web application in AngularJS
- Increased visibility of the company through more and more targeted posts on social media channels
- Used technologies: requirements management, stakeholder management, architecture decisions, NLP, LLM, Transformers, Python, FastAPI
DevOps
- Redesign and implementation of the complete infrastructure and CI/CD pipeline
- Technical infrastructure documentation and process description according to ISO/IEC 27001
- Stabilization and hardening of the infrastructure, as well as creation of a backup solution
- Elimination of platform downtime
- Deployment time accelerated by a factor of 4
- Increased time from ticket to production by a factor of 2 through dynamic review environments and JIRA automation
- Used technologies: Infrastructure as Code, Terraform, Hetzner, Cloud, Ansible, GitLab, CI/CD, Linux, Bash, Docker, Docker Swarm, Microservices, MariaDB, MinIO, Git
Interim CTO for decentralized cloud solution
- Overall responsibility for development and product strategy
- Introduction of structures and processes
- Planning of a new platform concept and support in implementation with Springboot, Java, and AngularJS
- Interface management
- Sharpened product strategy
- Lastingly improved interdisciplinary team collaboration, especially significantly increased the performance of the Indian team
- Introduced QA process to improve product quality
- Built new strategic partnerships and their technical integration into the platform
- Technologies used: project management, requirements management, stakeholder management, team development, architecture decisions, Agile, Clickup, Monday, Git, REST
Consultant and Data Scientist for insurance service providers
Mindhopper GmbH
- Gathering requirements, creating use and business cases
- Development and implementation of various AI services and automations for digitizing claims handling processes: document splitting, automated evaluation of receipts, document type recognition, claim type recognition, trade recognition, BelegGPT for asking questions about receipts
- Optimization of the internal claims review process by about 50 %
- Successful support in product placement at insurance companies
- Significant contribution to the new optimization of the insurance claims review process
- Technologies used: requirements management, stakeholder management, data analysis, prototype development, machine learning, LLM, NLP, image recognition, anomaly detection, text classification, GPT APIs, spacy, nltk, pytorch, scikit-learn, OpenCV, Python, FastAPI, REST, Docker, microservices, MongoDB, MySQL, Git, CI/CD
Consultant and Data Scientist for B2B trading platform
Windeit Software GmbH / Trusc GmbH
- Project management
- Development and implementation of AI-based business and use cases for a B2B trading platform for used hardware, e.g. price indication for used hardware, AI-based normalization of product specifications, intelligent product specification search for over 8 million products
- Improving platform UX by introducing AI-supported user input
- Creating new business cases for existing AI services
- Solution-oriented project management with 100% customer-focused results
- Technologies used: stakeholder management, data analysis, machine learning, NLP, LLM, spacy, nltk, pytorch, scikit-learn, OpenSearch, Python, FastAPI, REST, Docker, microservices, MariaDB, Git, CI/CD, JIRA, Confluence
Advanced Data Analysis & Optimization of BI dashboards
Windeit Software GmbH / Lenovo
- B2B trading platform*
Challenge: Inconsistent, uncleaned e-commerce data and not very meaningful dashboards made data-based decisions difficult.
Implementation:
- Data cleansing and ETL processes for consistent, quality-checked e-commerce data; implemented within the existing enterprise governance and permission structures
- Development of new, relevant KPIs for the trading platform and internal processes
- Optimization and expansion of the Power BI dashboards (usability, information density)
Success: Greater transparency in e-commerce processes and noticeably faster data-based decisions
Technologies used: Power BI, Python, Data Warehouse, ETL, PL/SQL
Development and Data Science
bridgingIT GmbH
- Feasibility project for a CV parser and a skill-matching algorithm
- Due to user acceptance and positive test results, the project led to the digitization and optimization of the recruiting process
- Technologies used: stakeholder management, requirements management, LLM, NLP, machine learning, Tensorflow, scikit-learn, spacy, Python, JavaScript, HTML, FastAPI, REST, Docker, SQLite, Git
Development and Data Science
- Development of a name generator for projects, products, and websites
- Creation of a database with all registered domain names worldwide (over 100 GB, 300 million records) to validate the availability of generated names
- Contribution to reducing external costs for naming workshops and agencies
- Technologies used: stakeholder management, data analysis, machine learning, Markov processes, NLP, scikit-learn, Python, FastAPI, GraphQL, REST, PostgreSQL, Docker, Git, CI/CD
Independent consultant for AI and digitalization
Freelancer
Design and implementation of a trading algorithm
Iron Eagle Capital GmbH
Challenge: For automated trading in currencies and crypto, an independent solution was to be developed that processes market data in real time, generates trading signals, and executes transactions automatically.
Implementation:
- Development and iterative optimization of a trading algorithm based on advanced time series analysis
- Implementation of the trading logic and platform integration with MQL5 and Python
Success: Development of a working end-to-end system for automated analysis of crypto and currency markets and execution of rule-based trading decisions in real time
Technologies used: MQL5, Python, time series analysis, WebSockets, algorithm development
Consultant and project management for a sales solution
Right Point GmbH
- Project management and product management
- Re-development of the AI-based sales solution
- Creation of innovative AI-based analyses for sales data (cross- and upselling potential, industry and customer analyses, lead potential analyses, efficiency and effectiveness analyses)
- Building a new development team and onboarding new developers
- Increasing the F1 score of the existing AI from 0.35 to 0.8
- Expanding the solution with additional key figures and metrics
- Speeding up the initial customer data analysis
- Creating new business cases for existing AI services
- Solution- and budget-oriented project management
- Technologies used: stakeholder management, data analysis, use and business case development, product management, architecture decisions, machine learning, time series analysis, anomaly detection, Tensorflow, scikit-learn, Python, MariaDB, Airflow, MLflow, Git, CI/CD
Head of the Competence Center for Self-Adaptive Systems
DAI-Labor Berlin
Data Scientist and Project Lead
DAI-Labor Berlin
Student Assistant in the Field of Machine Learning
TU Berlin
System Administrator
Humboldt Universität zu Berlin
Industry Experience
See where this freelancer has spent most of their professional time.
Experienced in Information Technology, Education, Professional Services, Insurance, Retail, and Banking and Finance.
Business Area Experience
See which departments and functions this freelancer has contributed to most.
Experienced in Information Technology, Project Management, Business Intelligence, Research and Development, Product Development, and Strategy.
Summary
Mathias Wilhelm helps companies unlock the economic potential of AI – through solid consulting, targeted implementation, and many years of experience in tech and business projects. With over 20 years of experience as a data scientist, developer, and project lead, he brings not only deep technical know-how, but also a strong understanding of business context and project strategy. Whether enterprise, start-up, or research – he combines entrepreneurial thinking with solid AI expertise.
Skills
Ai Consulting & Strategy: Development Of Use Cases, Feasibility Studies, Ai Roadmaps, Requirements Analysis
Data Science & Ml Engineering: Time Series Analysis, Anomaly Detection, Classification, Nlp, Ner, Document Analysis, Image Processing, Forecasting, Statistics, Data And Results Visualization
Prototyping & Product Development: From Idea To Mvp – Including Architecture, Development, Testing, Deployment, And Infrastructure
Devops/Mlops & Infrastructure: Ci/Cd, Mlops, Docker Swarm, Terraform, Ansible, Cloud (Aws, Hetzner, Etc.), Automation, Security
Technical Project Leadership: Steering Development Teams, Requirements Management, Quality Assurance
Industries: Sales, Insurance, Cloud, B2b Platforms, Recruiting, Finance, Automotive, Iot
Languages: Python, Sql, Bash, Javascript, Html/Css
Ml/Ds: Genai, Llm, Transformer, Image Recognition, Llm Prompt Engineering, Scikit-Learn, Tensorflow, Pytorch, Opencv, Pandas, Numpy, Mlflow
Databases: Mysql, Mariadb, Sqlite, Mongodb, Postgressql, Minio, S3
Frameworks: Spacy, Nltk, Pytorch, Tensorflow, Scikit-Learn, Opencv, Fastapi, N8n, Opensearch
Devops: Docker, Docker Swarm, Gitlab Ci/Cd, Terraform, Ansible, Linux, Git, Jira
Cloud/Hosting: Aws, Hetzner, Ovh, Cudo Computing, Vultr, Self-Managed Docker Swarm
Apis & Services: Api Design, Rest, Fastapi, Mqtt, Json, Xml
Languages
Education
TU Berlin
Doctorate (Dr.-Ing.), Ring-based multi-finger gesture control · Ring-based multi-finger gesture control · Berlin, Germany
TU Berlin
Dipl.-Ing. Computer Engineering · Computer Engineering · Berlin, Germany
Statistics
Experience
Global Experience
Expertise
Qualifications
Profile
Frequently asked questions
Have questions? Find more information here.
Daily Rate Distribution
The rates shown represent the typical market range for freelancers in this position based on recent contracts on our platform.
Actual rates may vary depending on seniority level, experience, skill specialization, project complexity, and engagement length.
Average rates for similar positions
Rates are based on recent contracts and do not include FRATCH margin.
The average daily rate is the mean of all daily rates from recent contracts of comparable freelancers on our platform.
The median daily rate is the middle value of all daily rates — half of comparable freelancers charge less, half charge more. Unlike the average, it is barely affected by outliers.
Actual rates may vary depending on seniority level, experience, skill specialization, project complexity, and engagement length.
Similar Freelancers
Discover other experts with similar qualifications and experience
Kai Kramer
Martin Musiol
Jürgen Fey
Tungi Dang
Arne Hendricks
Paul Webster
Eduard Van kleef
Yannick Schuchmann
Patrick Von der gönna
Jens Rehsack
Experts recently working on similar projects
Freelancers with hands-on experience in comparable project as a Implementation of an on-premise OCR solution with information extraction
Tobias Jaeuthe
Stephan Baier
Julien Look
Julian Balling
Michael Schiffer
Fadi Shoaa
Karl Estermann
Lino Giefer
Oliver Köhn
Jens Daube
Nearby freelancers
Professionals working in or nearby Berlin, Germany
Andreas Rühl
Chris Wolf
Björn Stark
Nune Isabekyan
Gilad Gotesman
Christian Pestel
Hanna Fischer
Anjana Rakesh
Patrick Hohensee
Anish Gupta
Most recent projects
FRATCH works with many companies and recruitment agencies. Here you will find our recently posted projects and opportunities.
