Consultant · Salesforce, Azure & AI · 4brands Reply
I build CRM solutions, data pipelines, and AI applications on Salesforce and Microsoft Azure, from field-sales automation to document and voice AI.
My background in applied mathematics shows up in how I work: benchmarks before model choices, test matrices before rollouts, and numbers I can back up.
Case studies from client and internal projects. Client and product names are changed or left out.
Two Salesforce apps that turn photographed drinks menus and expense receipts into CRM records. A vision model reads the document, deterministic code checks and matches it, and nothing is written until a person has reviewed it.
Every store visit gets a Perfect Store score: configurable checks, points per threshold, and a traffic light. I built the calculation as an Azure Data Factory and Azure SQL pipeline that reads from Salesforce and writes the results back reliably.
Sales users wanted their Power BI reports directly on Salesforce records, filtered to the customer they're looking at. I built an embedding where Apex brokers the Power BI token, so users never sign in twice and no secret reaches the browser.
Field reps have targets such as visits, distributed products, or custom KPIs. I moved the calculation of their actual values to Python Azure Functions and Azure SQL, orchestrated by Azure Data Factory, and wrote the results back to Salesforce.
When a field rep closes an in-store order, Salesforce now generates a branded PDF with both signatures and emails it to the right people. Getting it to actually arrive meant diagnosing SPF, DKIM, and DMARC.
A German voice assistant for field sales got noticeably worse after 10–20 minutes of use. As part of the project team, I worked on tracing the problem through production logs to a browser speech-state bug and on making the agent's CRM writes verifiable.
Set up a desktop AI system with vLLM behind an OpenAI-compatible gateway and measured throughput, tool calling, and fine-tuning, reaching 255 tokens/s at 100 parallel requests. A QLoRA fine-tune lowered accuracy, so I recommended the base model.
Backtested a risk model that decides which store visits get extra preparation time. With the same 20% capacity, it reached about 597 more at-risk visits than random selection (precision 88.5% vs. a 49.4% base rate, ROC-AUC 0.813).
Replaced a manually maintained Excel promotion calendar with a pipeline from the Fabric lakehouse through Power Automate to a VBA generator. It builds the calendar workbook automatically every weekday at 3:30.
Since July 2025, I’ve worked as a Consultant at 4brands Reply, following a working student role with the same team. My work combines CRM development, data integration, analytics, and AI.
Before that, I worked on language models at Diamant Software and researched multimodal data integration for industrial maintenance at Fraunhofer IPT. Alongside client work, I explore machine learning and optimization through my own projects.
Personal and university projects in machine learning, optimization, and software engineering.
A linear programming toolkit in Python: a two-phase Simplex solver with step-by-step tableaus, 2D/3D visualizations, sensitivity analysis, and a natural-language interface powered by an LLM.
A full-stack application for training and visualizing a linear regression model and a feedforward neural network, with a C++ engine, a Node.js backend, and a React frontend.
A nature-inspired metaheuristic for the Traveling Salesman Problem, implemented in Python for visualization and in C++ for performance, with a MATLAB animation of the search.
My work spans Salesforce applications, data pipelines, and AI services, from business workflows and API design to deployment, testing, and model evaluation.
Build CRM interfaces and workflows, connect external services, and validate business rules in Salesforce.
Build SQL-backed integration pipelines and Fabric lakehouse transformations, including incremental loads and data reconciliation.
Develop APIs and background workers with durable queues, retries, idempotency, and transactional data updates.
Integrate vision models into document workflows with structured extraction, schema validation, and human review. Apply a background in language models and multimodal RAG.
Set up local inference environments and benchmark answer quality, tool calling, throughput, and latency. Run QLoRA experiments and compare results against the base model.
Package and deploy API and worker services, build delivery pipelines, and verify changes with automated tests and operational checks.
Further technical foundations: applied mathematics and optimization, C++, MATLAB, pandas, NumPy, scikit-learn, data visualization, Elasticsearch, React, and HTML/CSS.
A strict JSON schema guarantees the shape of an AI extraction, not its truth, so this is how I design grounding, review and the write path around that gap.
A practical guide to the progress markers behind a Salesforce and Azure SQL round trip with Azure Data Factory, and why they are not interchangeable.
A pattern-level guide to embedding Power BI reports in Salesforce with an Apex token broker, a service principal, and row-level security that fails closed.
I’m happy to talk about CRM and data integration, applied AI, or a project you’re working on. Email is the fastest way to reach me.