Hire the Best Deep Learning Experts
in Armenia

7+ years building production computer-vision, OCR, edge-AI, and AI backend systems where accuracy, latency, and reliability matter. Typical work includes defect detection, live video/object tracking, on-device model optimization, document/OCR pipelines, OpenAI/RAG integrations, and launch rescue for AI products that need evaluation, observability, and clean API architecture. My strongest fit is a project with real data, a broken or uncertain pipeline, and measurable acceptance targets: accuracy, false positives, latency, edge-device constraints, cost, security, or launch readiness. Core work: - Computer vision / OCR: detection, segmentation, tracking, image analysis, OCR extraction, validation, reviewer workflows - Edge AI: ONNX/TensorRT conversion, model optimization, quantization-aware deployment, mobile/on-device inference - AI backend: OpenAI API, RAG, VLMs, source traces, confidence checks, FastAPI, PostgreSQL - Production delivery: tests, logging, monitoring, cloud deployment, handoff docs I am a good fit when AI output must be accurate, fast enough for production, reviewable, logged, and reliable enough for real business use.

I build credit scoring and loan default prediction models that help lenders approve more good borrowers, cut losses, and make underwriting decisions they can stand behind. I work in Python, and I focus on models that go into production not prototypes that get shelved. Before Upwork, I spent three years as an in-house Data Scientist at Ameriabank, one of Armenia's largest banks, building forecasting and risk models used in live operations. That experience shapes how I work: I care as much about how a model holds up over time and under review as I do about its accuracy on day one. I usually work with: — Fintech and alternative/SME lenders building or improving credit decisioning — Banks and credit unions that need scorecards built, rebuilt, or validated — Risk and analytics teams that need an extra pair of expert hands What I typically deliver: — Credit scoring and application/behavioral scorecards — Loan default prediction and PD models — PD / LGD / EAD modeling and IFRS 9 ECL work [ADD REAL PROJECT DETAIL] — Model validation, monitoring, and clear documentation — Risk segmentation and lending portfolio analytics — Fraud detection models for lending — Risk dashboards and SQL/Python pipelines that replace fragile Excel reporting Tools I rely on: Python (pandas, scikit-learn, statsmodels), XGBoost, LightGBM, logistic regression, and SQL, with FastAPI and Docker when something needs to be deployed. A few honest signals: Top Rated Plus, 100% Job Success, and $40K+ earned on Upwork. On a recent credit scoring project the client's feedback was simply "Highly recommended," and another noted my "amazing knowledge in ML and DS." [ADD REAL METRIC e.g. Gini/AUC, default-rate reduction, or approval-rate lift] How I work is straightforward: a short scoping call, then a clear plan with an honest estimate, then iterative delivery with documentation your team and any reviewer can actually follow. If something isn't the right fit for me, I'll tell you. Send a short message with your problem, your data, and your timeline. I usually reply within a few hours. The first 15-minute scoping call is free.

I am a full-stack Python, C++ AI/ML/ Computer vision / 3d reconstruction developer ✅ Top Rated PLUS Upwork Freelancer ✅ 15000+ hours worked ✅ 120+ Jobs Completed ✅ $300k+ earned Computer Vision and Machine learning - Computer Vision | Machine learning OpenCV,PCL,ROS,Detectron,YOLO, VTK, Intel Realsense, Zed camera, Zivid camera, NLP, Transformers - Computer Vision, Image Processing, OpenCV, OpenGL, MKL, ITK, VTK - Deep Learning, caffe, Tensor flow, Pytorch - YOLO, DETECTRON -Desktop application development using C++/Qt, Python -3d reconstruction, NLP using Matlab,R, Python, OpenCV, OpenGL,CUDA,OpenCL, MKL, ITK, VTK,PCL,ROS,R, Transformers. -Machine and Deep Learning using SVM, KNN, Neural Networks(TensorFlow, Yolo, Detection, Pytorch). -GUI development, sockets. -Stereo Vision and 3d reconstruction. SLAM and SFM algorithms implementation and improvement. -Video/Audio streaming over network using LIBVLC, FFMPEG, GSTREAMER. -Natural language processing using BERT, BART. -Development of technically complex projects and scientific articles. -Generic programming, OOP. -Complex algorithms & data structures.

✅ EXPERT-VETTED 🎖 TOP RATED PLUS (Top 1% 🥇 of GLOBAL talent on Upwork). I am a Data Scientist, AI, Machine Learning, and Software Engineer with over 9 years of professional experience building software at any scale. I used to work at large companies such as Yandex as well as in many successful startups. Now I am running Shapeion Technologies, providing software and ML consulting for companies around the globe, to accomplish their business goals. I have been managing startup projects with client teams and have helped draft product roadmaps, planning, and talent acquisition. I have a good network of engineer talent who can help build product MVPs effectively. Products made by me helped startups to raise millions of USD. I have a team of 15 professional Data Annotation specialists who have completed many projects in the Natural Language Processing and Computer Vision fields. They are fluent in the following data annotation platforms: CVAT, Doccano, and ProdiGy. Here is a brief summary of my expertise: ● 2000+ worked hours and 30+ projects completed PERSONALLY only at Upwork ● Created MVPs for various startups and companies that raised millions of USD ● Helped 30+ businesses to reach their goals in time I am proficient in: ✅ Data Science and Machine Learning: ● classification and regression tasks ● time-series prediction (ARIMA, ...) ✅ Natural Language Processing (NLP): ● Text classification (including Sentiment Analysis, Aspect-Based Sentiment Analysis) ● Named Entity Recognition (NER, automatic extraction of relevant entities, like names, locations, dates, etc) ● Chat-Bots (Deep Learning or Rule-Based chat-bots for various domains) Frameworks: SpaCy, HuggingFace, Transformers, NLTK, Langchain, Pinecone Annotation platforms: Doccano, ProdiGy, Clarifai Architectures: BERT, LongFormer, GPT-3, GPT-4 Domains: MedTech, E-Commerce, HRM, AD-tech ✅ Computer Vision (CV): ● Image Classification ● Object Detection (YoLo) ● Instance segmentation / Semantic Segmentation (Mask R-CNN) ● Optical Character Recognition (OCR) Frameworks: PyTorch, TensorFlow, Keras, DarkNet, Detectron, Detectron2, Fast.ai Annotation platforms: CVAT, Clarifai Domains: Medical Imaging, CCTV (surveillance), NFT, LegalTech (document analysis) ✅ Full-Stack Development: ● Backend (FastAPI, Flask, Django, django-rest-framework, Node.js, JavaScript, PHP, Yii Framework) ● Frontend (Vue.js, React.js) ● WordPress ● CI/CD (Docker, docker-compose, Kubernetes, k8s, Ansible, Terraform) ● Databases (PostgreSQL, MySql, MongoDB, OracleDB, ArangoDB, ClickHouse, GraphDB) ● AMQ (Celery, RabbitMQ) Technical skills: Python, Javascript, Go, Java, Spring Boot, Gradle, Node.js, Angular, Tensorflow, Keras, NumPy, Serverless Frameworks, Docker, Terraform, Ansible, Kubernetes, Jenkins, Spinnaker, GitOps, DataOps, AWS. GCP, Azure, JAMstack, GraphQL, MEAN Stack, React, Redux, ES6, Express, MongoDB, Redis, Linux, Etcd, Consul, Kinesis, Redshift, DynamoDB, Spark, Hadoop, Kafka, PostgreSQL, MySQL, ELK, Git, Flask, and Django. Thanks for your consideration! "Alexander was excellent. He managed expectations well and overperformed on the work. Communications were excellent. He saved us 100s of hours of work by automating our processes. He's more expensive than many people on Upwork, but appears to be exceptionally good an reliable." - Tom Mather

💼 Profile Overview 🚀 Top Rated | $1M+ Earned | 20K+ Hours | 160+ Successful Projects I’m a Senior Software Engineer & AR/VR Developer with over a decade of experience delivering high-performance, technically complex applications across multiple domains — from computer vision and deep learning to 3D simulation and cross-platform development. My clients include startups, research institutions, and enterprise teams looking for reliable, scalable, and innovative software solutions built with precision and performance in mind. 🧠 Core Expertise 💻 C / C++ / Python / Qt / Unity Highly optimized C++ development (TBB, MKL, BLAS, LAPACK, IPP) Multithreaded, performance-critical applications Complex algorithms and data structures Desktop apps using Qt (Windows, macOS, Linux) 🧩 Computer Vision & Deep Learning OpenCV, OpenGL, ITK, VTK, MKL DICOM image processing & medical imaging (ITK) YOLO, Detectron, TensorFlow, PyTorch, Caffe Automated data annotation pipelines 🕶️ AR / VR / XR Development Unity3D for mobile, desktop, and enterprise AR & VR apps for HoloLens 2, Oculus, and Mixed Reality platforms Integration of advanced vision models into immersive environments 📱 Cross-Platform & Mobile Integration Android & iOS development with native C++ plugin integration Video/Audio streaming using FFmpeg, GStreamer, LibVLC Real-time communication & networked systems ⚙️ Additional Skills SQA & UI Automation (Selenium) Scientific software & algorithm development Custom tools for data visualization, annotation, and analytics 🏆 What Clients Say ✅ “Delivered exactly what we needed — top-notch technical depth.” ✅ “One of the most professional developers we’ve ever worked with.” ✅ “Understood complex vision algorithms better than our in-house team.” 🔍 Why Choose Me? Proven Top Rated track record on Upwork Over $1M earned with 100% job success Trusted by clients for technical excellence and reliability Strong communication, fast delivery, and long-term project support 📩 Let’s Build Something Extraordinary Whether you need a vision-powered AR app, a high-performance C++ engine, or a deep learning-based system, I can help you take your idea from prototype to production with precision and performance. Let’s connect and discuss how I can bring your project to life.

I am a software engineer with strong mathematical background and professional experience since 2009. My expertise lies in developing advanced C++ and Python applications, particularly in the fields of machine learning, computer vision, data science, and mathematical optimization. What I Offer: * scientific(mathematical) research of your problem * technologies & market review * software design & development * online consultations * ability to work solo and as a part of a team * experience to manage small R&D teams * flexibility in project management techniques and communication approaches Solving difficult and non-standard problems via custom software development is what I like to do. Usually I deeply analyse a problem with respect to customer's needs and strive to propose solutions tailored to customer's specific context. At the same time, I understand the difference between must-haves and nice-to-haves and my primary goal is to satisfy basic requirements first like to fit into time schedule and budget and to reach minimal quality requirements. Details on the fields of my expertise: 1. Computer Vision: * image filtering and enhancing * localization and counting of objects in images * semantic/instance/panoptic segmentation of objects in images * anomaly detection in images * optical character recognition * object tracking/re-identification in video * novelty detection in video * camera calibration, bundle adjustment * 3D object reconstruction * structure from motion, SLAM * sensors fusion, trajectory smoothing Technologies: OpenCV, dlib, OpenMVS, OpenMVG, numpy, scipy, pandas, scikit-learn, scikit-image, TensorFlow, keras, PyTorch, HuggingFace 2. Machine Learning & Data Science: Below are the fields where I applied machine learning(including deep learning) and/or data science techniques: * computer vision * image/video processing * recommendation systems * NLP (limited experience) * trading (limited experience) As a ML-developer, I can characterize myself as: * familiar with neural networks, SVM, decision trees, ensemble learning, genetic algorithms and understanding the math behind them; * ready to apply statistical approaches and mathematical optimization techniques; * knowing pros and cons of manual feature engineering; * comprehending how size of dataset, complexity of model and quality of initial approximation affect result and amount of resources consumed by training procedure; * realizing significance of data augmentation techniques and always pedantic about selecting them for a given problem; * understanding the weaknesses of the backpropagation/gradient descent approach and how they impact deep learning architectures and the design of deep learning optimization algorithms; * widely using transfer learning, fine tuning, knowledge distillation and model quantization techniques; * aware of the data domain shift threat and doing everything possible to minimize its effects. Technologies: PyTorch, keras, TensorFlow, HuggingFace, scikit-learn, scipy, numpy, pandas. 3. Backend Development: I have solid experience in development of robust, efficient, and scalable backend systems in Python. Usually, I'm focused on designing and implementing a micro service architecture which supports high load computational tasks managed with the help of RESTful HTTP(S) or WS API. Technologies: docker, MySQL, PostgreSQL, SQLite, SQLAlchemy, MongoDB, redis, RabbitMQ, celery, FastAPI, tornado, flask, ELK. 4. Data structures and algorithms: * Designing data structures and algorithms with respect to problem statement, computational complexity, numerical precision, and hardware type. * Parallel computations. * Concurrent computations. * Distributed computations. * Micro services architecture for computational purposes. Technologies: C++ multithreading, OpenMP, MPI, Python multiprocessing, Python multithreading. 5. Numeric and Probabilistic Models, Mathematical Optimization: I have strong mathematical background and a Master degree in Applied Mathematics and Computer Science. That means I understand and can use the following fields of maths in my R&D: * analysis * linear algebra * graphs * probability and statistics * numerical methods * mathematical optimization methods Reasoning based on this fields may help to translate a business request into a solvable mathematical problem and to suggest some non-standard shortcuts to standard problems under specific conditions. If you have any scientific paper or thesis with mathematical models and want to reproduce its results, I can surely help you with that. Technologies: Jupyter notebook, numpy, scipy, pandas, PyTorch, gurobi, PuLP, eigen, CGAL, GSL, alglib, boost, ceres-solver etc. --------------------------------------------------------------------------- If you have any challenging problem related to my fields of expertise, please do not hesitate to contact me directly here. I will reply you shortly and try to suggest the most efficient way to handle it.