I build Retrieval-Augmented Generation and LLM systems that reach production, plus agentic AI, mobile apps and high-performance websites. Based in Turin, Italy, working globally.
From intelligent chatbots to warehouse management systems — every solution is custom-built, fast, and production-ready.
Intelligent chatbots that understand your customers, answer questions 24/7, and handle multilingual communication — powered by modern LLMs.
A smart assistant that searches your warehouse data in real time — checks stock, finds alternatives when items are unavailable, and answers complex queries instantly.
Custom mobile apps for iOS and Android — from personal productivity tools to business operations management, built with clean UX.
Fast, SEO-optimized websites that convert visitors into customers. Clean, modern design tuned for search rankings and real business results.
Give your team an AI assistant that reads your documents, manuals, PDFs and internal knowledge base — and answers questions with pinpoint accuracy.
Transform raw business data into clear dashboards and predictive models that inform smarter decisions — from sales forecasting to operational KPIs.
The AI and app development landscape is evolving fast. Here are some of the most in-demand solutions businesses are asking for right now.
Autonomous AI that performs multi-step tasks — booking, ordering, data entry — with human oversight where it matters.
Meet customers where they already are — assistants integrated with WhatsApp Business API.
Voice-first assistants for customer service, internal tooling, and hands-free warehouse operations.
Automated pipelines connecting CRM, email, inventory, and calendars — with AI at the center.
Four clinical platforms delivered as an independent consultant, plus the production AI I built at Gesan S.r.l.
An AI platform that analyzes medical PDFs, extracts clinical insights, and compares patient history using LLMs, RAG, and OCR — FastAPI, React, PostgreSQL, Docker.
A real-time infectious disease forecasting and environmental risk analytics platform running over Kafka, Spark, and data warehouse architectures.
Structured synthesis of historical clinical documentation for a research and development programme building a multi-tenant Hub and Spoke platform for pain-management care networks.
Developed and deployed a new API for WeLodge, a GDPR-compliant clinical document management system acting as a patient-level repository and registry.
A text and voice based lifestyle and medical chatbot for personalized guidance and more accessible medical support — for patients and clinical staff.
Fine-tuned BERT and Llama models with PyTorch and Hugging Face Transformers to classify medical reports and support faster, data-driven clinical decision-making.
NLP and OCR system to classify and route multilingual business emails to the correct departments in real time — cutting manual routing by 30%+.
Field and warehouse operations platform delivered as one codebase spanning API, web, and mobile.
I'm Amir Mokhtarian — an AI engineer and full-stack developer specializing in Retrieval-Augmented Generation and LLM systems that reach production, from retrieval architecture and evaluation through to deployment and monitoring. I architect agentic AI and data-warehouse-connected chatbots that let organizations query and act on their data through natural language, and I fine-tune transformer models (BERT, Llama) with PyTorch and Hugging Face for document classification, semantic search, and summarization. Eight years inside healthcare — from clinical training in medical imaging and anatomical sciences to daily hands-on work in PACS/DICOM, hospital information systems, and electronic patient records — means the AI I build is shaped around how clinical data and clinicians actually work.
Clients: Tab Technology, Gesan S.r.l., and ARTENSYS. Four clinical platforms across three signed engagements, all contracted through December 2026 — LexiCare's RAG layer and Wirgilio integration, the HealthTrace data layer, ReferOn for Cato Maior, and a new API for WeLodge. I own each engagement end to end and independently, from technical analysis and documentation through functional testing and production go-live.
Italian healthcare IT company, 30 years in market and over 100 employees, building digital solutions for hospitals, clinics, and primary care providers. Built LexiCare, a text and voice medical chatbot with STT/TTS and AWS Polly, fine-tuned BERT and Llama for medical report classification, and the HealthTrace forecasting platform over Kafka and Spark.
Operated imaging systems and worked daily inside PACS/DICOM workflows, hospital information systems, and electronic patient records — the clinical data infrastructure now underpinning the AI systems I build. Handled patient intake, scheduling, and reporting pipelines.
Coordinated a team of volunteers and managed the collection and distribution of donations, assigning tasks to ensure timely delivery to families in need.
Delivered CPR training sessions for community members, improving public knowledge of life-saving techniques.
Angela is a live RAG system, not a generic ChatGPT wrapper. Each question is embedded with Gemini, searched in PostgreSQL with both pgvector cosine similarity and full-text search, fused with reciprocal rank fusion, then filtered with maximal marginal relevance so the model only sees grounded chunks. Open the chat and ask — she answers from that knowledge base.
Custom AI chatbots using modern LLMs (GPT-4, Claude, Llama) — customer care bots, warehouse query assistants, voice-enabled interfaces, and WhatsApp Business integrations trained on your company's data.
Yes. Most projects connect to existing infrastructure — ERP, WMS, CRM, or databases — via APIs, without rebuilding from scratch. I design integrations that fit your current stack.
Absolutely. I work remotely with businesses worldwide. All project phases — discovery, delivery, and support — are managed entirely online. I communicate in English, Italian, and Persian.
A production-ready chatbot typically takes 4–8 weeks depending on complexity, data availability, and integration scope. Simple web chatbots can go live in 2–3 weeks.
React Native for cross-platform iOS and Android — one codebase, native performance, deployed to both app stores. Backend APIs are built with FastAPI or NestJS.
Have a project in mind? Tell me what you need and I'll get back to you as soon as I can.
You can also email me directly: mokhtarian2020@gmail.com
