# Writing > Technical notes on applied AI: automatic data extraction, AI call agents, internal process automation, embedded systems and machine learning in production. - URL: https://mojtabaamini.com/writing/index-en - Language: en - Author: Mojtaba Amini, AI Software Engineer (Verona, Italy) - Sep 2026 Document AI [Extracting data from PDF to Excel with AI: how it really works](https://mojtabaamini.com/writing/pdf-to-excel-ai-en) Retyping data from PDFs into Excel is the first job AI can take away. Here is how a reliable pipeline works, what changes between digital PDFs and scans, and how to try it right now. - Sep 2026 Embedded · IoT [From sensor to dashboard: how I design data acquisition for an IoT device](https://mojtabaamini.com/writing/sensor-to-dashboard-en) Before AI models there is data, and good data is designed starting from the sensor. The method I use, from firmware to logging to the dashboard, learned on monitoring devices and energy-storage systems. - Sep 2026 Embedded · Prototype [A glucose meter on Arduino: what an embedded prototype teaches](https://mojtabaamini.com/writing/arduino-glucose-meter-en) In 2020 I built a digital glucose meter prototype with an Arduino, an INA219 current sensor and an LCD. How it works, where the code was fragile, and why a prototype is not a medical device. - Sep 2026 Embedded · Aerospace [Telemetry for a CanSat: ground station, packets and checksums](https://mojtabaamini.com/writing/cansat-telemetry-en) In 2018 I wrote the telemetry ground station for a CanSat project for the AIAA competition: serial packets with checksums, flight instruments in JavaFX and a log file for every session. What I would do the same, and what not. - Aug 2026 AI Automatic Extraction [AI automatic data extraction: how to get clean data out of messy documents](https://mojtabaamini.com/writing/ai-automatic-data-extraction-en) A practical guide to automatic data extraction from PDFs, invoices, contracts and scans — the architecture that actually works in production, and the four mistakes that sink most projects. - Aug 2026 AI Call Agents [AI call agents: what they can really do for a business today](https://mojtabaamini.com/writing/ai-call-agents-en) Voice AI agents that answer the phone, qualify leads and book appointments are now genuinely production-ready. Here is what works, what does not, and how to deploy one without damaging your brand. - Jul 2026 Career · AI [From aerospace to AI: what turbulence taught me about machine learning](https://mojtabaamini.com/writing/aerospace-to-ai-en) I started in computational fluid dynamics at Sharif and ended up shipping production AI in Italy. The path is less strange than it sounds — both are about modelling systems too complex to solve exactly. - Jul 2026 Process Automation [Process automation is a mindset, not a tool](https://mojtabaamini.com/writing/process-automation-mindset-en) Every company wants "automation". Few can say which process, in what order, and how they will know it worked. Here is how I decide what to automate — across manufacturing, finance and enterprise .NET backends. - Jun 2026 Career · Team [Shipping AI across three countries and three languages](https://mojtabaamini.com/writing/multilingual-engineer-en) Iran, Germany, Italy. Persian, English, Italian. What building AI in three countries taught me about the part of the job that has nothing to do with code. - Jul 2026 Document AI · .NET [Automating document verification inside an enterprise .NET backend](https://mojtabaamini.com/writing/ai-document-verification-en) How I embed AI into enterprise .NET services at Studium Group — turning manual document review into an observable, reliable automated pipeline that scales. - Jun 2026 AI · Consulting [How AI actually improves internal company procedures (a practical guide)](https://mojtabaamini.com/writing/ai-internal-process-en) A field-tested playbook for using AI to improve internal company procedures — where to start, what to automate first, and how to measure the return. Lessons from consulting for manufacturing, finance and services SMEs in Italy. - Apr 2026 Reinforcement Learning [Why RL beats heuristics on the factory floor](https://mojtabaamini.com/writing/rl-beats-heuristics-en) A retrospective on replacing a rule-based scheduler at Dekali with an RL policy — and the silent rules nobody had written down. - Feb 2026 Document AI [Document AI is just plumbing. The plumbing is the product.](https://mojtabaamini.com/writing/document-ai-is-plumbing-en) Most production "AI" pipelines are 5% model and 95% taking heterogeneous PDFs, scans, Excel and rumours and turning them into a single source of truth. - Dec 2025 AI in SMEs [The first 90 days of an AI engineer in an Italian SME](https://mojtabaamini.com/writing/first-90-days-en) Translating "we should use AI" into a shipped feature people actually open. A field manual. - Nov 2025 Graph Neural Networks [A GNN that learned our Bill of Materials](https://mojtabaamini.com/writing/gnn-bom-en) How we encoded stairlift components as a graph and used message-passing to predict process time and delivery date in under ten minutes. - Oct 2025 LLM · RAG [A RAG pipeline that actually answers from the manual](https://mojtabaamini.com/writing/rag-that-works-en) Lessons from shipping a customer-support chatbot grounded in 1,400 pages of Extrema manuals — and reducing internal response times by 60%. - Aug 2025 Engineering [Reading CAN-bus from energy storage, without losing your mind](https://mojtabaamini.com/writing/can-bus-batteries-en) Notes from the Voltfang internship: parsing CAN frames, decoding DBCs, automating conformity certificates. - Jun 2025 Research · CFD [Adding LEM to CANTERA — a research engineer's rite of passage](https://mojtabaamini.com/writing/lem-cantera-en) How a Linear Eddy Model went into the open-source CANTERA library — across C++, Fortran and Python — and what that taught me about shipping into other people's codebases. - May 2025 .NET · Angular [How I expose Python models behind a .NET API consumed by Angular](https://mojtabaamini.com/writing/angular-dotnet-ai-en) The Dekali stack is .NET + Angular with Python models behind the curtain. Here is the wiring that keeps latency under 200 ms on the order list. - Mar 2025 RNN · Production [An RNN that learned the rhythm of an Italian factory](https://mojtabaamini.com/writing/rnn-factory-rhythm-en) Predicting which phase a job is in, and what it will cost by the time it ships, from telemetry no one had ever bothered to keep. - Jan 2025 DNN · Configurator [From spec to BOM in 90 seconds: the configurator that closes deals](https://mojtabaamini.com/writing/spec-to-bom-en) A deep neural net that turns customer technical specs into the optimal stairlift configuration, the full Bill of Materials and a delivery date the sales team can actually commit to. - Nov 2024 Document AI · Finance [Old vs New: designing "I don't know" into a financial report](https://mojtabaamini.com/writing/old-vs-new-idk-en) On the Adhoc SCF portfolio-analysis system, the most important UI affordance is not the number. It is the row that says "we cannot tell, please confirm". - Sep 2024 Product · Strategy [The CRM is the moat, not the AI](https://mojtabaamini.com/writing/crm-is-the-moat-en) Notes from a CRM modernisation in the Italian subsidised-finance sector. Why I keep recommending tiny AI features and bigger CRM investments. - Jul 2024 Data · Ethics [Scraping check24.de without becoming the villain](https://mojtabaamini.com/writing/scraping-without-villain-en) Notes from the GreenPocket data project: how to do large-scale tariff scraping ethically, legally, and without breaking the source site. - May 2024 AI in SMEs [Why AI PoCs die in week six, and how to ship in week eight instead](https://mojtabaamini.com/writing/sme-poc-curse-en) After three industrial projects and a dozen pitches, the pattern is the same. The PoC works in a notebook, dies on integration, and is buried by week ten.