Hello, I'm Kareem! I'm an AI Engineer specializing in Machine Learning, Deep Learning, NLP, and Computer Vision, dedicated to building scalable, data-driven solutions that drive real innovation. Whether it’s intelligent chatbots, automation, or advanced AI models, I craft solutions tailored to business needs—optimizing workflows, extracting deep insights, and enhancing efficiency with cutting-edge AI. - My Promise to you 🚀 Expert Problem Solver: From simple tasks to complex challenges, I deliver solutions that are not just functional but highly efficient and robust. 📚 Comprehensive Documentation: Receive thorough documentation covering AI models, algorithms, and architectures, ensuring you fully understand the process and outcomes. 💡 Clear Code, Clear Results: My code is clean, well-commented, and designed for maintainability, making it easy to follow and adapt. 💬 Responsive Support: Need assistance? I guarantee a same-day response, providing you with consistent support throughout the project. 🔍 Stable and Thoroughly Tested Solutions: I ensure that your solution is stable, well-tested, and handles edge cases with ease, minimizing any risk of issues. 🔄 Flexible and Agile Development: Requirement changes? No problem. I adapt swiftly, ensuring full cooperation and seamless adjustments as your project evolves. 💡 My Expertise 🔹 Core Skills ✔ Machine Learning & Deep Learning ✔ Natural Language Processing (NLP) ✔ Computer Vision & Image Processing ✔ Data Analysis & Visualization ✔ Web Scraping & Automation 🔹 Programming Languages ✔ Python | C++ 🔹 AI & ML Frameworks ✔ TensorFlow | PyTorch | Keras | Scikit-learn 🔹 Machine Learning & Deep Learning ✔ ML Models: Logistic Regression, Decision Trees, SVM, KNN, Ensemble Methods, PCA, Naive Bayes ✔ Deep Learning: CNNs, RNNs, Autoencoders, GANs, Transformers, Large Language Models (LLMs) 🔹 NLP & Chatbots ✔ Text Classification, Sentiment Analysis, Language Translation ✔ Chatbot Development (Llama 3, OpenAI) ✔ Summarization, Named Entity Recognition (NER) 🔹 Computer Vision ✔ Object Detection & Tracking | Image Classification | Image Segmentation ✔ Feature Extraction | Facial Recognition | Optical Character Recognition (OCR) 🔹 Web Scraping & Data Extraction ✔ Scrapy | BeautifulSoup | Selenium 🔹 Data Analysis & Databases ✔ Matplotlib | Seaborn | SQL Looking for an AI expert who brings expertise, professionalism, and a commitment to delivering scalable, high-impact solutions? Look no further! Contact me today to learn more about how I can help take your project to the next level.

$700K+ delivered on Upwork across 145 projects with a 100% Job Success Score. I architect and ship production AI systems, Voice AI agents handling live calls under 2-second latency, computer vision platforms processing 8M+ daily events, and full-stack products running in production for years. What I build: 𝗩𝗼𝗶𝗰𝗲 𝗔𝗜 & 𝗖𝗼𝗻𝘃𝗲𝗿𝘀𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝗔𝗴𝗲𝗻𝘁𝘀 Real-time voice pipelines combining speech-to-text, LLM reasoning, and natural voice synthesis. I built SloanConnect, an AI telephony platform that automates insurance verification calls end-to-end, navigates IVR systems, holds dynamic conversations with representatives, and scales to hundreds of concurrent calls. Sub-2-second response time, encrypted compliance handling, production-stable. 𝗖𝗼𝗺𝗽𝘂𝘁𝗲𝗿 𝗩𝗶𝘀𝗶𝗼𝗻 𝗮𝘁 𝗦𝗰𝗮𝗹𝗲 Lead AI Engineer on two production CV platforms: Adlytic - privacy-safe in-store retail analytics with GDPR-compliant pipelines, heatmaps, and POS integration, and SportsEye - real-time soccer analytics from broadcast feeds with 96% accuracy across player tracking, field mapping, and tactical metrics with zero on-field hardware. 𝗠𝗮𝗰𝗵𝗶𝗻𝗲 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 𝗦𝘆𝘀𝘁𝗲𝗺𝘀 Hybrid recommendation engines on PySpark processing 8M+ daily impressions. NLP and sentiment systems wired into live dashboards. Healthcare process automation with HIPAA-aligned data handling. Models built to live in real environments, not notebooks. 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗟𝗟𝗠 𝗪𝗼𝗿𝗸𝗳𝗹𝗼𝘄𝘀 (𝗖𝗹𝗮𝘂𝗱𝗲, 𝗢𝗽𝗲𝗻𝗔𝗜, 𝗺𝘂𝗹𝘁𝗶-𝗺𝗼𝗱𝗲𝗹) End-to-end AI agents shipped to production: a Shopify + Amazon SP-API customer service agent with order lookup, refunds, returnless-refund policies, Slack/Email approval gates, and Klaviyo sync. Multi-step LLM pipelines with tool use, RAG, function calling, and human-in-the-loop checkpoints. I design these as systems, not single prompts, and I'm comfortable across Claude, OpenAI, and open-weight models depending on what fits the workload. 𝗙𝘂𝗹𝗹-𝗦𝘁𝗮𝗰𝗸 𝗙𝗼𝘂𝗻𝗱𝗮𝘁𝗶𝗼𝗻𝘀 Every AI product needs scaffolding that holds. I've shipped Django/Node/React/Next.js platforms that ran 3+ years in production, Alchemist Real Estate, audIT.app, and a long-running B2B SaaS that grew from MVP to hundreds of users across $195K+ in contracted development. Distributed backends, REST/GraphQL APIs, AWS-native deployment, data pipelines, and the DevOps to keep them stable. 𝗛𝗼𝘄 𝗜 𝘄𝗼𝗿𝗸 As Lead Engineering Architect, I stay hands-on through the whole lifecycle, discovery, architecture, core implementation, deployment, and the long tail. I don't hand off and disappear. Most of my Upwork work is repeat business or long contracts measured in years, not weeks. If you're building a Voice AI agent, a computer vision platform, an LLM-driven workflow, or a full-stack AI product that needs to survive real load - let's talk.

𝐄𝐱𝐩𝐞𝐫𝐭-𝐕𝐞𝐭𝐭𝐞𝐝: 𝐓𝐨𝐩 𝟏% 𝐨𝐧 𝐔𝐩𝐰𝐨𝐫𝐤 🥇 Senior Agentic AI Developer with 6+ Years of Experience | Shipped 50+ Role-Critical AI Agent + RAG Chatbots | Conversational AI | Senior AI First Full Stack Developer with Python(FastAPI + React) | Principal AI/ML | GenAI, LLM, Fine-tuning, Cloud(Azure, AWS, GCP) | I'm Samran Elahi, a Senior AI / Machine Learning Engineer who has worked extensively with NLP and Agentic AI. I'm equally skilled in automating custom workflows, developing LLMs, building RAG Powered Chatbots, and designing Machine Learning Models along with fine-tuning and Deploying them on AWS / Azure / GCP, or Runpod. ✅ 𝐖𝐡𝐚𝐭 𝐈 𝐁𝐫𝐢𝐧𝐠 𝐭𝐨 𝐓𝐚𝐛𝐥𝐞 — 𝐀𝐈 & 𝐌𝐚𝐜𝐡𝐢𝐧𝐞 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐦𝐞𝐧𝐭 𝐒𝐞𝐫𝐯𝐢𝐜𝐞𝐬 𝐰𝐢𝐭𝐡 𝗙𝘂𝗹𝗹 𝗦𝘁𝗮𝗰𝗸 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗺𝗲𝗻𝘁 𝗘𝘅𝗽𝗲𝗿𝘁𝗶𝘀𝗲 ■ Natural Language Processing Expertise ✓ Chatbots and Conversational AI Development ■ AI Agents + RAG Chatbots Development Expertise ✓ Custom AI Agents built with LangChain / LangGraph. ✓ Role-critical MCP Agents for complex task handling. ✓ RAG-powered Chatbots with integrated LLMs (OpenAI, Claude, Mistral.) ✓ Frontend: React, Streamlit, Chainlit | Backend: FastAPI, Django. ✓ Vector DBs: Pinecone, FAISS, Weaviate. ■ Conversational AI with Real-Time Communication Capabilities ✓ Tools: ElevenLabs, Voice.ai, Retell.ai, Vapi integrations. ✓ Real-time API integration for seamless voice interactions. ■ Custom ML Model Development, Integration, and Deployment ✓ Supervised Learning: Regression, Classification. ✓ Unsupervised Learning: Clustering, Dimensionality Reduction, Anomaly Detection. ✓ Hyperparameter Tuning: GridSearch, RandomSearch, Bayesian Optimisation. ✓ Parameter-Efficient Fine-Tuning (PEFT). ✓ Model Deployment: Runpod, AWS, GCP, EC2. ✓ Model & Data Versioning: DVC. ✓ REST API Development: FastAPI, Flask, Django. 📢 Available, let's have a quick chat to go over the scope of your project. Best, Samran Elahi

I am a Machine Learning Engineer with four years of experience working across deep learning research, large scale AI systems, and production model deployment. Over the years, I have worked extensively in medical imaging, computer vision, NLP, signal processing, and large language models, building systems that range from experimental research pipelines to deployed real world AI applications. My day to day work primarily involves Python, PyTorch, TensorFlow, Keras, HuggingFace Transformers, sentence transformers, scikit learn, OpenCV, Pandas, and NumPy. I enjoy working deeply on both the research and engineering sides of machine learning, especially problems that require understanding model behavior rather than simply applying existing architectures blindly. A large part of my background is research driven. I have authored multiple peer reviewed publications in indexed journals and IEEE conferences, including publications in Neurocomputing, Healthcare Analytics, Engineering Applications of Artificial Intelligence, Telematics and Informatics Reports, and other Elsevier and IEEE venues. My research has focused heavily on explainable AI, healthcare AI, and advanced deep learning systems. Some of my published work includes CARDxnosis, an explainable knowledge driven framework for ECG diagnosis and clinical report generation, an explainable AI system for trustworthy arrhythmia detection, a CNN RNN Attention hybrid architecture for automatic modulation classification, ensemble deep learning approaches for lung cancer detection from CT scans, and SRGAN based white blood cell image generation and classification pipelines. Alongside published work, I am currently involved in research on brain tumor segmentation, ADHD and ASD classification from brain connectome graphs, epileptic seizure prediction from EEG signals, and interpretable tabular learning using graph neural networks combined with Kolmogorov Arnold Networks. Beyond research, I have substantial hands on experience building and deploying production grade AI systems. One of my major recent projects was LaborBERT v4, a domain adaptive transformer fine tuning system processing hundreds of thousands of records through a large scale training pipeline. The project involved multiple experimental setups including contrastive learning, masked language model pretraining, temporal contrastive learning, cross attention based fusion, multi task training, and Matryoshka Representation Learning. I have also built hybrid embeddings plus LLM systems for taxonomy mapping using OpenAI embeddings alongside locally hosted LLaMA and Mistral models through Ollama. In addition, I have worked on deployed clinical AI systems and a portable on device diagnostic AI solution with embedded deep learning models for point of care inference, which gave me valuable experience in optimization, deployment constraints, inference design, and production reliability. My broader project portfolio includes vehicle detection using Mask R CNN, human activity recognition on the Kinetics 700 dataset, facial keypoint detection with MultiRes UNet, semantic segmentation pipeline redesign, Stable Diffusion based image editing workflows, toxic comment classification, RASA based conversational AI systems, and large scale scraping and automation pipelines using Playwright and Selenium. I have also worked with Flask and Django based deployment pipelines and cloud hosted ML systems. From an engineering perspective, I care strongly about clean and maintainable systems. I follow disciplined workflows involving modular code design, Git based version control, reproducible experimentation, structured evaluation, bootstrap validated metrics, and detailed documentation. I am also comfortable preparing scientific reports, research papers, and journal submissions using both LaTeX and Word. What ties all of this together is that I genuinely enjoy solving difficult technical problems, especially the kind that require balancing research depth with practical engineering constraints. I am most motivated by projects where thoughtful experimentation, careful system design, and real world usability matter equally.

LiveKit, WebRTC, SIP/VoIP, Electron, AI, video, audio, real-time software development at scale. 600M+ AI call minutes a month (Nucleus). $10M+ ARR EdTech (BrainCert). Enterprise VMS surveillance for 50,000 users (VALT). 𝗙𝗿𝗲𝗲 𝗣𝗥𝗗 𝗳𝗼𝗿 𝗻𝗲𝘄 𝗽𝗿𝗼𝗷𝗲𝗰𝘁𝘀, 𝗳𝗿𝗲𝗲 𝗰𝗼𝗱𝗲 𝗿𝗲𝘃𝗶𝗲𝘄 𝘄𝗶𝘁𝗵 𝗶𝘀𝘀𝘂𝗲 𝗿𝗼𝗼𝘁𝘀 𝗼𝗿 ~𝟱𝟬𝗵 𝘁𝗿𝗶𝗮𝗹 - 𝗳𝗼𝗿 𝗲𝘅𝗶𝘀𝘁𝗶𝗻𝗴 𝗼𝗻𝗲𝘀. I'm the CEO of Fora Soft, and personally architect every project. You get an owner accountable for delivery and senior judgment from day one, not a junior learning WebRTC on your budget. ⚠️ DM or invite me - I respond within 2 hours. WHAT I'VE BUILT (selected): → SpeedSpace (LiveKit) - remote video production on LiveKit, WebRTC, and FFmpeg. Makes content for Netflix, HBO, and Electronic Arts. Shown at Paris Fashion Week. → Nucleus .com (SIP/VoIP) - WebRTC and SIP business phone system for the Canadian telecom Fibernetics. 600M+ call minutes a month, 5,000+ businesses, AI voice agents, SOC II, HIPAA. → FRP.live (Electron) - desktop and AI music platform. 720,000+ licensed tracks from Sony, Universal, and Virgin. Used by 12,000+ DJs. → VocalViews .com (WebRTC) - research platform with 1M+ participants, used by Samsung, McDonald's, Google, Netflix, TikTok, Adobe. → Tyxit .com (Electron, audio) - Electron desktop app for sub-30ms live music. Performed at the Montreux Jazz Festival. $1.3M raised, partnered with Karmine Corp. → VALT ipivs .com (Wowza, WebRTC) - video recording, streaming, surveillance SaaS platform. 770+ US organizations, 2,500 IP cameras, 50K+ users. From $1M revenue in the launch year to $9.7M now. 12 years’ partnership. → Live Eye (Electron) - AI retail surveillance. 10K+ locations, $5M+ fraud prevented. → BrainCert .com - the first HTML5 + WebRTC virtual classroom. 100K+ customers, 500M+ classroom mins. → Sprii - €365M+ in live commerce sales. 72K+ live events. → StreamLayer - interactive sports overlays, $14.1M raised. NBC, CBS, Red Bull, Chelsea FC. PROBLEMS I SOLVE (with proof from real projects): • When third-party SaaS gets too expensive at scale: replaced GoogleDocs, Zoom and Discord with a custom LiveKit platform, 2,000 students in one live class (Scholarly: 15,000+ users) • Sub-30ms latency for live music (Tyxit, Electron app: Montreux Jazz) • SIP/VoIP calling at scale: WebRTC-to-SIP bridges, FreeSWITCH, Telnyx (Nucleus, 5K+ businesses) • AI surveillance on Electron desktop apps: $5M+ fraud prevented across 10K+ locations (Live Eye) • AI voice agents processing 600M+ call min/mo (Nucleus) • Secure video conferencing at enterprise scale: up to 500 participants, AES-256 (iMind: used by PwC and national telecoms) • Computer vision: 99.5% face recognition, 500K+ plates/day (Mindbox .ai) • Live commerce at €365M+ sales (Sprii: 72K+ events, 3K+ brands) • Sports tech: Premier League (Vokalo), NBC/CBS/Red Bull (StreamLayer, $14.1M raised) SPEC-FIRST AGENTIC AI: I work spec-first, not vibe-first. As CEO and architect, I write the full spec and architecture before any code. Then my senior team drives AI agents against strict guardrails to build production-grade systems. You get AI speed with the stability of a senior team. TECH I WORK WITH DAILY: ◆ Real-time, video, and audio: WebRTC, LiveKit, MediaSoup, MOQ, Ant, Kurento, Janus, Jitsi, Agora, FFmpeg, GStreamer, Wowza, HLS/DASH, RTMP/RTSP ◆ VoIP and telecom: SIP, FreeSWITCH, Telnyx, Twilio, Vonage ◆ Backend: Node.js, NestJS, Python, Go, PHP, PostgreSQL, MySQL, MongoDB, Redis ◆ Frontend and desktop: Electron, React, Next.js, Vue, Angular, JavaScript, TypeScript ◆ AI/ML: OpenAI API, Whisper, Deepgram, Speechmatics, Amazon Polly, TensorFlow, PyTorch, OpenCV ◆ Mobile: React Native, Flutter, Swift, Kotlin ◆ Infra: AWS, GCP, Azure, Docker, Kubernetes, CI/CD WHO I HELP: • Startups building their first video MVP • Scale-ups hitting WebRTC performance walls • Enterprises needing HIPAA/SOC2 video HOW I WORK: 1. 30-min call – assess your project honestly (including if we're NOT the right fit) 2. Written architecture + estimate within 24 hours 3. Dedicated team assigned within 1 week 4. I stay involved throughout – architecture reviews, tech decisions, escalations WAYS TO ENGAGE: build from scratch, drop in on the hard WebRTC/LiveKit/SIP/Electron modules, or rescue a stalled build. WHY FORA SOFT: 100% JSS, $10M+ earned, 925 jobs, 403K+ hours, 400+ clients across 17+ countries, 20 years in video, audio, and real-time software, Top WebRTC Developer 2022 and Top Telecom Developer 2024, 1 of 400 developers passes our internal selection. CLIENT FEEDBACK: "We've been working together for 10 years and continue. Fora Soft planned and built our entire video SaaS from scratch across web, iOS and Android." - Dustin, CTO @ VALT (50k users, 770 organizations, 2,500 IP cameras) ⚠️ Press Invite or message me. I'll send back a PRD, architecture, and an estimate within 24 hours.

I specialized in AI Agents, RAG systems, workflow automation, and enterprise AI applications. I have built 40+ AI solutions including knowledge assistants, document intelligence platforms, multi-agent systems, AI chatbots, and business process automation tools using OpenAI, Claude, Gemini, LangGraph, FastAPI, and vector databases. 🔑 What that means for you: Autonomous AI agents that replace repetitive human tasks. RAG chatbots that answer questions over your actual documents, databases, and internal tools and not hallucinated guesses LLM pipelines (GPT-4o, Claude, Gemini) integrated cleanly into your existing backend, and no rip-and-replace n8n / Make automation workflows that connect your CRM, Slack, email, and APIs into one intelligent system Production-grade backends (FastAPI, Django) on AWS / Azure / GCP that scale when you grow Top Rated Plus · 100% Job Success · 5-star average 💡 WHAT I BUILD FOR YOU ▸ AI Agents & Agentic Workflows Multi-step LangGraph agents that plan, reason, and execute across tools and APIs. Built for SaaS products, internal ops, and customer-facing automation. Not toy demos — production pipelines with error handling, retries, and logging. ▸ RAG-Based Chatbots & Knowledge Assistants Your documents, your database, your knowledge — made conversational. I build RAG systems using Pinecone, Weaviate, or pgvector + OpenAI/Gemini that return accurate, cited answers over thousands of files. ▸ LLM Integrations & Prompt Engineering GPT-4o, Claude 3.5, Gemini — function calling, structured outputs, tool use, reasoning chains. I handle the full stack from prompt design to API integration to output validation. ▸ Workflow Automation (n8n, Make, Zapier) Trigger-based and agentic workflows connecting your CRMs, databases, communication tools, and APIs. Cut operational overhead without replacing your team. ▸ Backend for AI Products Async FastAPI / Django services, background task queues, cloud storage, auth systems — designed to support AI-heavy workloads at scale on AWS, Azure, or GCP. ▸ Voice & Generative AI (Advanced) Real-time voice AI pipelines, voice cloning APIs, talking avatar systems, LLM-powered report generation — for teams building the next generation of AI-native products. 🔧 MY TECH STACK AI/LLM: OpenAI · Claude · Gemini · LangChain · LangGraph · llamaindex · CrewAI RAG: Pinecone · Weaviate · pgvector · FAISS · ChromaDB Automation: n8n · Make · Zapier · Webhooks Backend: FastAPI · Django · Flask · Celery · Redis Cloud: AWS · Azure · GCP · Docker · GitHub Actions Databases: PostgreSQL · MongoDB · MySQL · Supabase 📋 HOW WE WORK TOGETHER Discovery Call (30 min) — I understand your workflow, pain points, and current stack Scoped Proposal — You get a clear breakdown of deliverables, timeline, and cost Build & Iterate — Weekly check-ins, live demos, clean commits. You always know the status Deploy & Hand Off — Production deployment + documentation + 2 weeks of post-launch support I don't disappear after delivery. I build systems you can maintain, extend, and trust. Ready to replace manual work with intelligent AI? 📩 Message me now — I respond within 4 hours and offer a free 15-minute scope call for serious projects.

🏆Top Rated for more than 10 years on Upwork | 🏆 70+ projects | 🏆 100% Job Success Score | 🏆 16 years experience | 🏆Working with 3x enterprise companies and so many start-ups/individual clients Production-grade computer vision systems for real-world environments — edge deployment, system integration, and measurable business outcomes. I graduated as an Electrical Engineer M.Sc. (Embedded Software Developer faculty). I am an expert in AI/Machine learning (object detection) and IoT/Edge Computing (Raspberry Pi and Jetson Orin NX/Nano, Deepstream 6.x, 7.x, Triton/TAO toolkit). I prefer long-term projects, but you can hire me for shorter ones also as a consultant. Strength: - Computer Vision/Machine learning Project architect/design - technical stack, cost of long-term usage, refine user requirement etc. - Python Machine learning projects (Tensorflow, Keras, PyTorch) - especially with computer vision, object detection (YOLO), image segmentation and pose tracking (OpenPose), shape recognition (dlib) and action recognition (X3D, mmaction etc.) - Develop/maintain/upgrade your existing Deepstream/Triton based application - Embedded image processing solutions (Edge computing): computer vision on Jetson Orin (NX/Nano) with Deepstream7.0 (earlier version too 6.x/7.x, Triton, TAO toolkit) or on Raspberry Pi, using Intel Movidius Neural Compute Stick (NCS) + OpenVINO. Train a model for specific purpose, provident dataset collection/preparation/preprocessing task management, I specialized for object detection (up to YOLOv12) - AWS, GCP (Google Cloud Platform) VM inference and training setup and execution/deployment - NVIDIA GPU experience: Tesla T4, L4, A100, H100 GPU for training and inference/server production. - Home automation/Smart home - lighting, building engineering, heating/cooling, Home assistant: integrate Amazon Alexa, Phillips Hue, Sonos, Apple TV, Google Nest, custom solutions - Raspberry Pi (Linux - Raspbian) - programming and teaching in Python (RS232, RS485, TCP communication), Wi-Fi settings, Home assistant Services: - Short (30-60 minutes) consultation - AI Architect tasks: responsible for the architecture of AI solutions, which includes planning, implementing, and managing AI technologies within an organization or project - I can speak English and Hungarian (magyar) and a little bit Spanish (castellano, espanol) AI Computer Vision | MLOPS | Edge Computing | AI Automation | AI strategy | AI consultant

𝐓𝐨𝐩 𝐑𝐚𝐭𝐞𝐝 𝐏𝐥𝐮𝐬 𝐫𝐞𝐚𝐥‑𝐭𝐢𝐦𝐞 𝐜𝐨𝐦𝐦𝐮𝐧𝐢𝐜𝐚𝐭𝐢𝐨𝐧𝐬 𝐞𝐧𝐠𝐢𝐧𝐞𝐞𝐫 (𝐖𝐞𝐛𝐑𝐓𝐂/𝐋𝐢𝐯𝐞𝐊𝐢𝐭 + 𝐒𝐈𝐏/𝐕𝐨𝐈𝐏 + 𝐓𝐰𝐢𝐥𝐢𝐨/𝐀𝐬𝐭𝐞𝐫𝐢𝐬𝐤). 𝐈 𝐛𝐮𝐢𝐥𝐝 𝐥𝐨𝐰‑𝐥𝐚𝐭𝐞𝐧𝐜𝐲 𝐚𝐮𝐝𝐢𝐨/𝐯𝐢𝐝𝐞𝐨, 𝐬𝐜𝐚𝐥𝐚𝐛𝐥𝐞 𝐬𝐭𝐫𝐞𝐚𝐦𝐢𝐧𝐠, 𝐫𝐞𝐜𝐨𝐫𝐝𝐢𝐧𝐠/𝐞𝐠𝐫𝐞𝐬𝐬 𝐩𝐢𝐩𝐞𝐥𝐢𝐧𝐞𝐬, 𝐚𝐧𝐝 𝐕𝐨𝐢𝐜𝐞 𝐀𝐈 (𝐒𝐓𝐓/𝐓𝐓𝐒 + 𝐜𝐨𝐧𝐯𝐞𝐫𝐬𝐚𝐭𝐢𝐨𝐧𝐚𝐥 𝐀𝐈) 𝐭𝐡𝐚𝐭 𝐰𝐨𝐫𝐤𝐬 𝐢𝐧 𝐩𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧. I build enterprise-grade real-time systems, 𝐖𝐞𝐛𝐑𝐓𝐂 / 𝐋𝐢𝐯𝐞𝐊𝐢𝐭, 𝐒𝐈𝐏/𝐕𝐨𝐈𝐏 𝐭𝐞𝐥𝐞𝐩𝐡𝐨𝐧𝐲, and 𝐕𝐨𝐢𝐜𝐞 𝐀𝐈 (STT/TTS + conversational AI). If you're dealing with 𝐥𝐚𝐭𝐞𝐧𝐜𝐲, 𝐜𝐚𝐥𝐥 𝐝𝐫𝐨𝐩𝐬, 𝐍𝐀𝐓/𝐓𝐔𝐑𝐍 𝐢𝐬𝐬𝐮𝐞𝐬, 𝐫𝐞𝐜𝐨𝐫𝐝𝐢𝐧𝐠 𝐟𝐚𝐢𝐥𝐮𝐫𝐞𝐬, 𝐬𝐜𝐚𝐥𝐢𝐧𝐠 𝐩𝐫𝐨𝐛𝐥𝐞𝐦𝐬, or 𝐕𝐨𝐢𝐜𝐞 𝐀𝐈 𝐢𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐢𝐨𝐧, I can help you deliver a stable solution end-to-end. I'm 𝐓𝐨𝐩 𝐑𝐚𝐭𝐞𝐝 𝐏𝐥𝐮𝐬 on Upwork with strong long-term delivery history, and I focus on what clients actually care about, 𝐫𝐞𝐥𝐢𝐚𝐛𝐢𝐥𝐢𝐭𝐲, 𝐬𝐜𝐚𝐥𝐚𝐛𝐢𝐥𝐢𝐭𝐲, and 𝐦𝐞𝐚𝐬𝐮𝐫𝐚𝐛𝐥𝐞 𝐩𝐞𝐫𝐟𝐨𝐫𝐦𝐚𝐧𝐜𝐞 (join time, RTT, jitter, packet loss, concurrency, CPU/bandwidth cost). I'm also an inventor on a 𝐠𝐫𝐚𝐧𝐭𝐞𝐝 𝐔𝐒 𝐩𝐚𝐭𝐞𝐧𝐭 (𝟏𝟐,𝟒𝟓𝟐,𝟑𝟏𝟕) in teleconferencing/real-time communication, so RTC is not just "a feature" for me; it's the core. 𝐖𝐡𝐚𝐭 𝐈 𝐝𝐨 𝐛𝐞𝐬𝐭 ✅ 𝐖𝐞𝐛𝐑𝐓𝐂 / 𝐋𝐢𝐯𝐞𝐊𝐢𝐭 (𝐥𝐨𝐰‑𝐥𝐚𝐭𝐞𝐧𝐜𝐲 𝐚𝐮𝐝𝐢𝐨/𝐯𝐢𝐝𝐞𝐨 + 𝐬𝐜𝐚𝐥𝐢𝐧𝐠) • LiveKit architecture (self-hosted) + production hardening • SFU scaling strategy (regions, routing, concurrency planning) • TURN/STUN/ICE reliability for strict NAT / enterprise networks • Performance tuning: jitter, packet loss, bitrate adaptation, simulcast/SVC strategy • Recording/egress pipelines + media routing + stream health monitoring ✅ 𝐓𝐞𝐥𝐞𝐩𝐡𝐨𝐧𝐲 (𝐒𝐈𝐏/𝐕𝐨𝐈𝐏) + 𝐓𝐰𝐢𝐥𝐢𝐨/𝐏𝐥𝐢𝐯𝐨 + 𝐀𝐬𝐭𝐞𝐫𝐢𝐬𝐤 • SIP trunking integrations (inbound/outbound), routing, call control • Asterisk: IVR, queues, dialplans, conferencing, PBX integrations • Twilio Programmable Voice: webhooks, recordings, call flows, media streaming • WebRTC ↔ SIP bridging (browser/app calling into PSTN) ✅ 𝐕𝐨𝐢𝐜𝐞 𝐀𝐈 (𝐒𝐓𝐓/𝐓𝐓𝐒 + 𝐂𝐨𝐧𝐯𝐞𝐫𝐬𝐚𝐭𝐢𝐨𝐧𝐚𝐥 𝐀𝐈) 𝐭𝐡𝐚𝐭 𝐰𝐨𝐫𝐤𝐬 𝐨𝐧 𝐫𝐞𝐚𝐥 𝐜𝐚𝐥𝐥𝐬 • Real-time STT streaming (partials, low latency) + speaker-aware workflows when possible • Streaming TTS with interruption handling (barge‑in) and turn-taking logic • Voice agent orchestration (including Vapi-style patterns) •Call summaries, structured data extraction, CRM updates, analytics • Guardrails + fallbacks (human handoff, escalation, retry logic) ✅ 𝐀𝐈/𝐌𝐋 + 𝐂𝐨𝐦𝐩𝐮𝐭𝐞𝐫 𝐕𝐢𝐬𝐢𝐨𝐧 (𝐰𝐡𝐞𝐧 𝐢𝐭 𝐬𝐮𝐩𝐩𝐨𝐫𝐭𝐬 𝐭𝐡𝐞 𝐩𝐫𝐨𝐝𝐮𝐜𝐭) I integrate AI/ML in a production-first way that doesn’t break real-time constraints: • OCR + computer vision pipelines (live overlays, detection, automation) • ML/LLM components for classification, routing, QA scoring, intent detection • PyTorch/TensorFlow integration when custom models are required • Generative AI features (RAG / knowledge-based responses) with accuracy controls 𝐇𝐨𝐰 𝐈 𝐰𝐨𝐫𝐤 (𝐬𝐨 𝐩𝐫𝐨𝐣𝐞𝐜𝐭𝐬 𝐝𝐨𝐧'𝐭 𝐝𝐫𝐢𝐟𝐭) 𝟏. 𝐐𝐮𝐢𝐜𝐤 𝐝𝐢𝐬𝐜𝐨𝐯𝐞𝐫𝐲: goals, target platforms, expected concurrency, latency KPI, constraints 𝟐. 𝐏𝐥𝐚𝐧: architecture + milestones + risks (written, clear scope) 𝟑. 𝐁𝐮𝐢𝐥𝐝 & 𝐢𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐞: clean implementation + reviews + docs 𝟒. 𝐏𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧 𝐫𝐞𝐚𝐝𝐢𝐧𝐞𝐬𝐬: monitoring, alerts, load testing, handover 𝐓𝐲𝐩𝐢𝐜𝐚𝐥 𝐝𝐞𝐥𝐢𝐯𝐞𝐫𝐚𝐛𝐥𝐞𝐬 • Working, deployed solution (Docker/Ubuntu) • Reliable call/stream flows + error handling • Recording/transcription + storage integration if needed • Monitoring dashboards + alert rules • Documentation + runbooks + next-step roadmap 𝐖𝐞𝐛𝐑𝐓𝐂, 𝐋𝐢𝐯𝐞𝐊𝐢𝐭, 𝐕𝐨𝐢𝐜𝐞 𝐀𝐈, 𝐒𝐈𝐏, 𝐕𝐨𝐈𝐏, 𝐓𝐰𝐢𝐥𝐢𝐨, 𝐏𝐥𝐢𝐯𝐨, 𝐀𝐬𝐭𝐞𝐫𝐢𝐬𝐤, 𝐓𝐞𝐥𝐞𝐩𝐡𝐨𝐧𝐲, 𝐒𝐓𝐓, 𝐓𝐓𝐒, 𝐋𝐋𝐌, 𝐂𝐨𝐧𝐯𝐞𝐫𝐬𝐚𝐭𝐢𝐨𝐧𝐚𝐥 𝐀𝐈, 𝐕𝐚𝐩𝐢, 𝐃𝐞𝐯𝐎𝐩𝐬, 𝐃𝐨𝐜𝐤𝐞𝐫, 𝐔𝐛𝐮𝐧𝐭𝐮, 𝐅𝐅𝐦𝐩𝐞𝐠, 𝐎𝐁𝐒 𝐒𝐭𝐮𝐝𝐢𝐨, 𝐎𝐦𝐧𝐢𝐯𝐞𝐫𝐬𝐞, 𝐀𝐈/𝐌𝐋, 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐯𝐞 𝐀𝐈, 𝐀𝐫𝐭𝐢𝐟𝐢𝐜𝐢𝐚𝐥 𝐈𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐜𝐞, 𝐌𝐚𝐜𝐡𝐢𝐧𝐞 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠, 𝐃𝐞𝐞𝐩 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠, 𝐀𝐥𝐠𝐨𝐫𝐢𝐭𝐡𝐦 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐦𝐞𝐧𝐭, 𝐂𝐨𝐦𝐩𝐮𝐭𝐞𝐫 𝐕𝐢𝐬𝐢𝐨𝐧, 𝐎𝐂𝐑, 𝐏𝐲𝐓𝐨𝐫𝐜𝐡, 𝐓𝐞𝐧𝐬𝐨𝐫𝐅𝐥𝐨𝐰, 𝐆𝐨𝐥𝐚𝐧𝐠, 𝐀𝐯𝐚𝐭𝐚𝐫, 𝐃𝐢𝐠𝐢𝐭𝐚𝐥 𝐓𝐰𝐢𝐧, 𝐀𝐈 𝐂𝐡𝐚𝐭𝐛𝐨𝐭.

I'm an AI engineer specializing in AI agent development, RAG systems, generative AI, and LLM automation — $90K+ earned on Upwork, 27 projects delivered, Top Rated, with production AI and enterprise software running for clients across healthcare, SaaS, and compliance-heavy industries. Most clients come to me after an AI prototype that never made it to production. I build AI systems that hold up under real users, real data, and real compliance requirements — not demos. 🤖 AI Agent Development Multi-agent systems and autonomous workflows using LangGraph, CrewAI, AutoGen, and custom agent architectures. Agents that make decisions, call your APIs, process documents, and complete real business tasks end to end. 📚RAG Systems & Enterprise Document Intelligence Retrieval Augmented Generation pipelines over your documents and knowledge bases — semantic search, chat-with-your-data, and enterprise document intelligence. I've built RAG systems searching 100,000+ documents with sub-second latency using Qdrant, Pinecone, ChromaDB, and FAISS, including fully on-premise, air-gapped deployments for compliance-heavy organizations. 🧠 LLM Integration & AI App Development Production integration of OpenAI API (GPT-4/GPT-5), Anthropic Claude, Gemini, Llama, and open-source models — including hybrid model routing that sends simple tasks to fast, cheap models and complex reasoning to heavy ones, cutting inference costs without cutting quality. AI chatbots, document processing, classification, and AI features inside existing products. ⚙️ AI Automation & Workflows LLM-powered automation using n8n, Make, and Python — lead processing, data extraction, content pipelines, and internal tooling that removes manual work. If your team is copy-pasting between tools, I can probably automate it. ☁️Cloud Infrastructure & AI Security AWS (EC2, ECS, S3, Bedrock), GCP, Docker, Kubernetes — scalable deployments, GPU inference, and AI security done properly: encrypted data pipelines, on-premise and air-gapped LLM deployments, role-based access control, and audit trails that satisfy enterprise compliance. Recent work includes enterprise software builds like a document intelligence platform with on-premise LLMs and SAP integration, and a multi-agent healthcare SaaS platform aggregating data from 32 medical directories with LLM-powered analysis and content generation. How I work: ✅ Production-ready code with testing, documentation, and monitoring ✅ Fixed-price milestones or clearly scoped hourly engagements ✅ Clear async communication and regular progress demos ✅ Ongoing optimization after deployment — most clients retain me after the first build Stack: Python, FastAPI, LangChain, LangGraph, OpenAI API, Claude, Qdrant, Pinecone, PostgreSQL, Supabase, n8n, Make, AWS, GCP, Docker, Kubernetes, Custom Software Development, Highlevel Development, Claude Opus, Claude Fable, n8n custom deployment, Enterprise Software, AI Automation Message me with what you're trying to build or automate — I'll respond with specific questions about your use case, not a template pitch.

I build production-ready Generative AI systems - image generation pipelines, AI video, LLM apps, and autonomous agents. Not just prototypes that work in a demo, but systems that ship, scale, and keep running in production. Most AI projects don't fail at the idea stage - they stall at 80%. The prototype works, the demo impresses, and then it never quite becomes a reliable product clients can actually ship. That last stretch - consistency, structured outputs, error handling, real deployment - is exactly what I do. 10+ years building software, now focused entirely on Generative AI and Agentic AI. 𝐖𝐡𝐚𝐭 𝐈 𝐁𝐮𝐢𝐥𝐝: • AI image generation: text-to-image pipelines using ComfyUI, SDXL, Flux, and Stable Diffusion with custom workflow orchestration and production-grade quality control. • AI video generation and editing: text-to-video pipelines, AI-powered editing, clip generation from long-form content, and motion-controlled animation systems. • LoRA and model fine-tuning: DreamBooth training, PEFT/LoRA adapters for domain-specific image generation, automated dataset curation, and quality scoring. • AI character and digital persona creation: identity-consistent character generation across hundreds of outputs with automated quality gating. • LLM integration & orchestration: GPT-4, Claude, Llama, and Mistral integrated into business systems with structured outputs, schema validation, and production-safe response handling. • RAG Sytems: document Q&A, enterprise knowledge bases, semantic search using LangChain, LlamaIndex, and vector databases (Pinecone, Chroma, Weaviate, FAISS). • AI agents & automation: autonomous multi-step agents using CrewAI, AutoGen, MCP integrations, and n8n workflows with LLM-powered decision-making. • AI SaaS products & copilots: end-to-end generative AI platforms from architecture to deployment, including user interfaces, API layers, and scalable cloud infrastructure. • Prompt engineering: systematic prompt design, A/B testing, and output evaluation for consistent, high-quality results across text, image, and video, using tools like Promptfoo and Langfuse. 𝐅𝐫𝐨𝐦 𝐏𝐫𝐨𝐭𝐨𝐭𝐲𝐩𝐞 𝐭𝐨 𝐏𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧: Building the AI is half the job - getting it live and stable is the other half. I handle the full deployment so your project doesn't stall at "it works on my machine": CI/CD for ML pipelines, model versioning, containerized inference (Docker), monitoring, and cost-efficient GPU provisioning on RunPod, Modal, Lambda Labs, and major clouds (AWS, GCP, Azure). You get a working product, not just a prototype. 𝐑𝐞𝐜𝐞𝐧𝐭 𝐖𝐨𝐫𝐤: ✔️ 𝗗𝗔𝗧𝗔𝗦𝗜𝗠: AI character generation platform with dual ComfyUI pipelines (SDXL + Z-Image), automated LoRA training, quality gating system, and GPU queue serialization. Processing thousands of images with identity-locked consistency. ✔️ 𝗗𝗼𝗰𝘂𝗺𝗲𝗻𝘁𝗟𝗟𝗠: Multi-document RAG chat, automated comparison, canvas workflows, and presentation generation from uploaded files. ✔️ 𝗙𝗹𝗼𝘄𝗠𝗶𝗻𝗱: Built an AI-powered automation system using GPT-4, n8n, and FastAPI to handle data evaluation, scoring, and intelligent routing.Eliminated manual workflows, improved efficiency, and delivered with a 5.0-star client rating. ✔️ 𝗛𝗶𝗿𝗲𝗛𝗮𝗿𝗿𝘆: AI agent platform backend with multi-agent orchestration, LLM-powered task automation for marketing, content generation, and business analytics, with production-safe structured outputs and deterministic state transitions. I hold a 100% Job Success Score. If your project involves generative AI, image or video generation, LLM integration, fine-tuned models, agentic systems - or you need any of it actually deployed and running - send me the details and I'll tell you exactly how I'd build it. 𝐓𝐞𝐜𝐡: ComfyUI, SDXL, Flux, LoRA, Stable Diffusion, DreamBooth, PyTorch, OpenCV, GPT-4, Claude, Llama, LangChain, LlamaIndex, Neo4j, CrewAI, AutoGen, OpenClaw, Pydantic AI, Microsoft Semantic Kernel, NotebookLM, MCP, n8n, RAG, Pinecone, Chroma, FAISS, Runware, Promptfoo, Langfuse, Python, FastAPI, APIs and Webhooks, Supabase, Convex DB, Docker, Kubernetes, AWS, GCP, MS Azure, Oracle Cloud, RunPod, Modal, Fal.ai, Local GPUs based Infrastructure and its configurations. Note: I don't take on NSFW, adult, or explicit content projects.