Hey there, I'm
Taimir Aguacil
Head of IIoT & AI
Bridging deep tech (ETH Zürich) and business strategy (IMD) to architect scalable AI platforms and build high-performance teams that deliver measurable ROI.
Is This You?
I'm deliberate with my time and generous with it when there's real signal.
I've operated at every scale.
What I've learned is that the hardest problems are rarely technical. They're about clarity of vision, trust between people, and the courage to move when it's uncomfortable. I stay open to conversations because that's how I keep learning, and occasionally, it's how the best things start.
If any of this sounds like you, my inbox is open.
Building the Future One Platform at a Time
I help businesses turn complex technology into commercial growth. My career spans from co-founding a Series A marketplace with over a million users, to currently leading a 60-person AI and IIoT department for a €4B+ industrial giant. Ultimately, I connect the dots between engineering, product, and strategy, empowering cross-functional teams to build scalable platforms that deliver measurable ROI.
Core Expertise
Product & Business Strategy
Vision to execution
AI & Machine Learning
Platforms, not proofs of concept
Team Leadership
Cross-functional, multi-country
Stakeholder Alignment
Executive communication
"Without commitment, you will never start, but without consistency you'll never finish"
Professional Experience
Endress+Hauser Group
Head of IIoT & AIDefining global technology vision for IIoT & AI, driving the convergence of Cloud, Edge, and AI into secure, scalable platforms powering the Netilion ecosystem. Restructured a 60-person department toward a product-oriented vision, drove 20% YoY revenue growth, and own the IIoT cybersecurity posture (Zero-Trust, IT/OT alignment). Launched an Agentic AI solution projected to save €9M annually. Member of the Group CIO’s core AI team shaping company-wide AI strategy.
Quantum Dice
Advisor | Product Strategy, Solution Architecture, AI & Quantum SystemsFollowing this Oxford University spin-off focused on cybersecurity innovation from inception. Bringing expertise from telecom, hardware, and embedded systems to shape next-gen products. Architecting the 'Quantum Entropy as a Service' (QaaS) offering and advising on go-to-market strategy for probabilistic computing.
Kallam Technologies
Advisor | Product Strategy, AI Architecture, Legal TechShaping the product vision and technical architecture for an AI-powered document intelligence platform purpose-built for legal disputes. Designed the hybrid RAG search engine and multilingual OCR pipeline, and helped define the 'Virtual Intern' autonomous research agent. The platform transforms weeks of manual document triage into minutes, 60% faster review, 100% source traceability.
Mazadat, Smarkt Holding
Co-Founder, CTOCo-founded a recommerce platform. Led end-to-end technology and product strategy, building the platform from scratch to over 1 million downloads. Deployed ML models for content moderation and anomaly detection, cutting manual review by 70%.
UCRAFT
Product Management Consultant | Product-Led Growth, Pricing Strategy, AnalyticsEngaged with this Silicon Valley website builder to sharpen their product-led growth engine. Mapped critical user paths and introduced predictive analytics to eliminate onboarding bottlenecks, achieving a 70% decrease in onboarding time and a 50% increase in B2B conversion rates.
Trapeze Switzerland
Product Lead & Sr. Software EngineerArchitected a next-gen IoV router that drove $20M in presales and 80% market adoption in year one. Optimized embedded systems to boot 75% faster. Co-led R&D on AI-powered passenger counting and emergency detection for autonomous vehicles.
ETH Zürich
Research EngineerDesigned signal processing and ML algorithms for mobile IoT transceivers, outperforming state-of-the-art by 20%. Built a GPS tracking system using compressive sensing that reduced data transmission by 40%. Published in Springer Nature (High Dimensional Optimization and Probability: With a view towards Data Science).
Featured Projects
Research & OSS
Deep tech · LLM · FinTech
Consulting
Strategy · Growth · Analytics
Enterprise & Startups
AI platforms · IIoT · Series A
Spearheaded an Agentic RAG platform to resolve a critical workforce capacity crisis. The platform provides internal support engineers with a secure AI...
Spearheaded an Agentic RAG platform to resolve a critical workforce capacity crisis. The platform provides internal support engineers with a secure AI-powered chat interface to query tens of thousands of proprietary technical documents and receive cited, context-aware answers within seconds. Projected to deliver €9M in annual savings by automating 70% of repetitive support cases with break-even in under 12 months.
Architecture: Multi-model agentic architecture combining OCR, knowledge graphs (LightRAG), vision-language models, and agentic reasoning. GDPR-compliant via private Azure EU tenant.
Mazadat Platform
Series A eCommerce MarketplaceCo-founded and scaled the first X2C online-offline auctioning marketplace in EMEA, taking it from a blank canvas to a Series A company. Built and led ...
Co-founded and scaled the first X2C online-offline auctioning marketplace in EMEA, taking it from a blank canvas to a Series A company. Built and led a distributed engineering team of 17 across 5 countries, establishing the culture and processes that enabled rapid iteration without sacrificing quality. When early metrics revealed conversion bottlenecks, I orchestrated a strategic product pivot that doubled engagement time and drove a 30% boost in retention. Beyond the platform itself, I pioneered the region's first buy-now-pay-later integration and deployed ML systems for fraud detection and content moderation that cut manual review by 70%. Achieved 1 million+ downloads and 100,000+ MAU in under 12 months.
Co-supervised development of a multi-modal AI safety platform for autonomous transit, combining NLP-based speech intelligence with computer vision for...
Co-supervised development of a multi-modal AI safety platform for autonomous transit, combining NLP-based speech intelligence with computer vision for real-time passenger monitoring, emergency detection, and occupancy analysis. Enables proactive intervention and compliance with emerging autonomous transit safety regulations.
Served as principal architect from inception to industrialization for a next-generation 4G+ router (10Gbit/s) designed to solve complex legacy integra...
Served as principal architect from inception to industrialization for a next-generation 4G+ router (10Gbit/s) designed to solve complex legacy integration challenges in public transit. The solution introduced core network acceleration and high-performance packet processing to enable reliable vehicle-to-everything communications at scale. I defined the client integration roadmap and led stakeholder demos that translated technical capabilities into commercial traction,facilitating $20M+ in presales and achieving 80% market adoption within the first year.
Stack: Yocto Linux, OpenStack, Open vSwitch, OpenDaylight, DPDK/ODP for high-performance packet processing.
Co-developed the core engine powering Quantum Dice's probabilistic computing stack,a new computational paradigm positioned between classical and quant...
Co-developed the core engine powering Quantum Dice's probabilistic computing stack,a new computational paradigm positioned between classical and quantum computing. Orbit tackles NP-hard combinatorial optimization problems that are intractable for traditional approaches, using probabilistic bits (p-bits) and replica-based sampling with simulated annealing. Beyond simulation, the platform is architected with a clear path toward hardware deployment, bridging the gap between research validation and commercial product. Currently demonstrating performance on real-world problems like Max-Cut and portfolio optimization to validate market potential with enterprise customers.
Built an AI-powered document intelligence platform purpose-built for the unique demands of legal disputes. Kallam automates the most time-consuming pa...
Built an AI-powered document intelligence platform purpose-built for the unique demands of legal disputes. Kallam automates the most time-consuming parts of case preparation: reading thousands of documents, extracting dates and entities, splitting bundled PDFs, classifying exhibits, and generating fully-cited chronologies,work that traditionally takes weeks of associate time. The platform features a 'Virtual Intern' autonomous research agent that can conduct multi-turn legal analysis across case files, hybrid semantic-lexical search with reranking, and a Word Add-In for seamless citation management. Multilingual OCR and 100% source traceability ensure nothing gets lost in translation or attribution.
Impact: 60% faster document review in real arbitration cases with zero training required.
Delivered a GenAI strategy engagement for one of Europe's largest energy companies, defining practical applications of LLMs and agent...
Delivered a GenAI strategy engagement for one of Europe's largest energy companies, defining practical applications of LLMs and agentic AI for Building Information Modelling (BIM) workflows. The engagement covered document intelligence for technical archives, automated compliance checking, and knowledge retrieval systems,translating cutting-edge AI capabilities into actionable roadmaps aligned with enterprise constraints.
Built and deployed a full-stack AI SaaS platform designed to handle document intelligence at scale. Instrumentum transforms large volumes of unstructu...
Built and deployed a full-stack AI SaaS platform designed to handle document intelligence at scale. Instrumentum transforms large volumes of unstructured documents into an interactive, searchable knowledge base,enabling users to get instant, cited answers instead of manually reviewing hundreds of pages. The platform orchestrates multiple frontier LLMs (Claude, GPT-5, Gemini) through a unified agentic interface, with hybrid retrieval combining semantic and graph-based RAG for precision at volume. Cloud drive integrations (SharePoint, OneDrive) and multi-language support make it enterprise-ready from day one.
Tech: Multi-model agentic chat, hybrid retrieval pipeline, Supabase, Stripe billing, multi-language support.
Designed and built the delivery infrastructure for Quantum Dice's Entropy as a Service offering,enabling organizations to access certified qua...
Designed and built the delivery infrastructure for Quantum Dice's Entropy as a Service offering,enabling organizations to access certified quantum-generated randomness at scale. The platform serves high-stakes applications in cryptography, financial modeling, and scientific computing where the integrity of random numbers is critical. Built for enterprise reliability with per-client bandwidth management, usage tracking, and seamless integration via Python and C/C++ client libraries.
Developed an NLP-based classification model to automatically categorize articles for CWire's contextual advertising platform. Designed the feature eng...
Developed an NLP-based classification model to automatically categorize articles for CWire's contextual advertising platform. Designed the feature engineering pipeline and trained multi-label classifiers to map content to IAB taxonomy categories, enabling precise ad targeting at scale.
Built an open-source AI-powered portfolio management system that combines rigorous financial analytics with natural language interaction. The platform...
Built an open-source AI-powered portfolio management system that combines rigorous financial analytics with natural language interaction. The platform features a multi-agent architecture where an orchestrator routes queries to specialized agents for transaction handling, portfolio analysis, and market research. Supports stocks, ETFs, bonds, crypto, and multi-currency portfolios with comprehensive metrics including Sharpe ratio, VaR, alpha/beta analysis, and Monte Carlo simulations. Includes portfolio optimization using Hierarchical Risk Parity and Mean-Variance methods, along with what-if scenario modeling.
Exposes 30+ tools via Model Context Protocol (MCP),actively used with Claude to manage my personal portfolio.
Led an IMD consulting engagement with CEAT (India's leading tire manufacturer) to tackle declining call center conversion rates across 40,000+...
Led an IMD consulting engagement with CEAT (India's leading tire manufacturer) to tackle declining call center conversion rates across 40,000+ monthly outbound calls. The project combined strategic analysis with hands-on AI development: we analyzed call recordings to identify conversion leaks and success patterns, built customer persona mappings, conducted competitive feature analysis, and derived the 'golden rules' that separate successful calls from lost sales. On the technical side, I architected a real-time Dynamic Prompt Engine using multi-agent orchestration (sentiment analysis, persona matching, product expertise, competitive intelligence) that delivers contextual sales guidance to agents in 4-8 seconds.
Impact: Projected $25M business uplift opportunity with 2.9x ROI and payback in under 2 years.
LLM Benchmarking Platform
Load Testing & Performance Comparison for LLM APIsComprehensive load testing and benchmarking platform for LLM API endpoints. Enables performance comparison across OpenAI, Azure OpenAI, and Anthropic ...
Comprehensive load testing and benchmarking platform for LLM API endpoints. Enables performance comparison across OpenAI, Azure OpenAI, and Anthropic using LiteLLM as unified proxy gateway. Features Flask dashboard for real-time monitoring and Locust for concurrent load testing with custom prompt scenarios.
Extended Vercel's AI Chatbot into a production-grade platform. Integrated 30+ LLM providers (Claude, GPT, Grok, Gemini, DeepSeek, Lla...
Extended Vercel's AI Chatbot into a production-grade platform. Integrated 30+ LLM providers (Claude, GPT, Grok, Gemini, DeepSeek, Llama, Mistral, Perplexity). Designed for power users who need model flexibility, document grounding, and artifact creation in a single unified interface.
Features: Document RAG with Upstash Vector, custom AI agent builder, Python sandbox (Pyodide), image generation (DALL-E).
Developed a novel location tracking system for IoT devices that fundamentally rethinks how GPS data is captured and transmitted. Traditional approache...
Developed a novel location tracking system for IoT devices that fundamentally rethinks how GPS data is captured and transmitted. Traditional approaches require constant sampling, draining battery life and generating massive data volumes. BckTrck applies compressive sensing theory to prove that accurate full-track reconstruction is possible from dramatically fewer samples. I implemented two reconstruction algorithms,a LASSO regression approach offering flexibility without prior training, and a neural network approach optimized for speed on mobile AI accelerators. The system also handles the practical challenges of GPS: signal noise, urban canyon effects, and dropped data points. Published in Springer Nature volume on High Dimensional Optimization.
Achieves accurate full-track reconstruction from just 20% of original samples, reducing data transmission by 40% while improving accuracy over naive full-sampling in noisy conditions.
Designed the signal processing algorithms powering high-speed 4G transceivers for vehicle-to-everything (V2X) and IoT communications. Built a complete...
Designed the signal processing algorithms powering high-speed 4G transceivers for vehicle-to-everything (V2X) and IoT communications. Built a complete simulation framework from the ground up covering the full transceiver chain: time and frequency synchronization, channel estimation, and equalization. The algorithms achieved 100% ETSI compliance while outperforming existing state-of-the-art solutions by 20% on key metrics. Progressed the work from MATLAB prototyping through high-level synthesis to FPGA implementation, bridging the gap between algorithmic research and silicon-ready designs.
This foundational research directly informed the architecture of Trapeze's GPR3 4G+ router for IoV communications.
Academic Background
Beyond Business
"I never heard that name before, where is it from?", my father invented it by replacing letters in a name my mother suggested. The closest similarity is the Taymyr Peninsula in Siberia, which I have yet to visit.
I live in Switzerland with my beautiful French wife. I speak 4 languages and have heritage from 4 countries. Off the clock, you’ll find me kitesurfing in the Red Sea, training for my next trail ultra, or skiing in the Alps. I’m happiest when the stakes are real.
Let's Get in Touch
I've been blessed to work with brilliant people throughout my journey. Whether you're interested in extreme sports or product management, I'd love to connect.