Hire the Best Automatic Speech Recognition Freelancers
in Bangladesh

Need annotation done right the first time with zero quality issues? That's exactly what I deliver. I'm a Data Annotation Specialist with hands-on experience in audio/speech, image/video, and sentiment & intent labeling plus a Quality Assurance background that means I don't just label data, I verify it meets production-ready standards. I hold a C2 (Proficient) level in English, making me especially effective on English-language NLP, transcription, and sentiment projects where linguistic nuance directly impacts model accuracy. My 100% Job Success Score on Upwork reflects the consistency and reliability clients come back for. What sets me apart from other annotators: ✅ QA-Trained Eye — I've worked as a Quality Analyst on annotation projects, reviewing labeled datasets for consistency, guideline compliance, and edge-case errors before they reach your model ✅ Project Handling Experience — I've managed annotation workflows end-to-end, not just labeled tasks — so I understand timelines, batching, and delivery pipelines from a lead perspective ✅ C2 English Proficiency — Ideal for transcription, audio labeling, sentiment tagging, intent classification, and any task where English accuracy is non-negotiable ✅ 100% Job Success Score — Every client. Every project. Delivered right. ✅ Platform Fluency — Labelbox, Scale AI / Appen, Excel & Google Sheets Annotation types I cover: Audio & Speech — Transcription, speaker diarization, noise/silence tagging, ASR dataset labeling Image & Video — Bounding boxes, polygon segmentation, object classification, frame labeling Sentiment & Intent — Emotion tagging, intent classification, NLP dataset preparation, chatbot training data My promise to you: Clean data. Met deadlines. Clear communication. And a QA mindset that catches problems before they cost you time and money. 📩 Tell me about your dataset — I'll tell you exactly how I can help.

I am an expert AI/ML Engineer with hands-on expertise in Machine learning, Deep Learning, Computer Vision, NLP, Generative AI, and LLMs. I leverage advanced techniques to deliver innovative and impactful AI solutions for your needs. Areas of expertise: ✅ Machine Learning (Data Analysis and Visualization, Classification, Regression models, Predicting with Supervised, Unsupervised Techniques) ✅ Deep Learning (Deep Neural Networks, Transfer Learning, fine-tuning models, XAI, optimization, exploring models) ✅ Computer Vision and Image Processing (Object detection/segmentation/tracking, Image/Video classification, Image generation, OCR) ✅ NLP (Text classification, Topic modeling, NER, Semantic similarity, Text generation, Question Answering) ✅ Generative AI (Fine-tuning LLMs, RAG, Prompt Engineering, embeddings, vector database, Chatbots and AI Assistants) ✅ Audio Data Analysis (Audio classification, Sound segmentation, TTS, STT) Tech Stack: ✅ ML/DL tools: TensorFlow, Keras, PyTorch, Scikit-Learn, XGBoost, LightGBM, CatBoost ✅ NLP: LangChain, OpenAI(GPT-4, Embeddings, ChatGPT), Local LLMs, Huggingface, Gemini, NLTK, Textblob ✅ Computer Vision: Ultralytics, OpenCV, Pillow, Scikit Image, YOLO, Detectron2, OpenPose ✅ Data Analysis/Visualization Tools: Pandas, NumPy, Matplotlib, Seaborn, Plotly, Altair ✅ Cloud services: AWS, Google Cloud, Azure ✅ Project Management and Version Control: Git, GitHub ✅ MLOps: MLFlow, Docker, Github actions, Airflow, DVC ✅ Deployment: StreamLit, Flask, FastAPI ✅ Languages: Python My commitment to excellence and continuous learning ensures that I stay at the forefront of AI/ML advancements. Let’s connect to explore how I can contribute to your AI/ML projects and help you achieve your business goals with cutting-edge solutions.

Your AI project needs to 𝘄𝗼𝗿𝗸 𝗶𝗻 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻 — on real hardware, with real user data, at a cost that makes business sense. Not just a prototype. A system your users can actually rely on. That's what I focus on. My recent work includes 𝗿𝘂𝗻𝗻𝗶𝗻𝗴 𝟰.𝟯𝗚𝗕 𝗼𝗳 𝗔𝗜 𝗺𝗼𝗱𝗲𝗹𝘀 𝗲𝗻𝘁𝗶𝗿𝗲𝗹𝘆 𝗼𝗳𝗳𝗹𝗶𝗻𝗲 on a $200 Android phone by dissecting Google AI models, 𝗰𝘂𝘁𝘁𝗶𝗻𝗴 𝘃𝗼𝗶𝗰𝗲 𝗽𝗿𝗼𝗰𝗲𝘀𝘀𝗶𝗻𝗴 𝗰𝗼𝘀𝘁𝘀 𝗯𝘆 𝟵𝟴% through local voice model deployment, and publishing one of the 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻 𝗠𝗖𝗣 𝘀𝗲𝗿𝘃𝗲𝗿𝘀 𝗼𝗻 𝗡𝗣𝗠. The engineering is public — my GitHub and NPM packages are there for you to review. If your AI project needs to 𝘄𝗼𝗿𝗸 𝗼𝗳𝗳𝗹𝗶𝗻𝗲, 𝗽𝗿𝗼𝘁𝗲𝗰𝘁 𝘂𝘀𝗲𝗿 𝗽𝗿𝗶𝘃𝗮𝗰𝘆, 𝗼𝗿 𝘀𝗰𝗮𝗹𝗲 𝘄𝗶𝘁𝗵𝗼𝘂𝘁 𝗿𝘂𝗻𝗮𝘄𝗮𝘆 𝗔𝗣𝗜 𝗰𝗼𝘀𝘁𝘀 — I'd be glad to help you figure out the right approach. ✅𝗪𝗛𝗬 𝗖𝗟𝗜𝗘𝗡𝗧𝗦 𝗖𝗛𝗢𝗢𝗦𝗘 𝗧𝗢 𝗪𝗢𝗥𝗞 𝗪𝗜𝗧𝗛 𝗠𝗘 ✔️ I've Deployed AI Where No Cloud Reaches I engineered a Kotlin Android application running 4.3GB of LLM, Vision, and Audio models entirely offline on 4GB RAM devices — no cloud dependency. No data leaves the device. ✅What this means for you: If your use case demands on-device AI, offline capability, or strict data privacy — this is a problem I've already solved and shipped. ✔️ I Can Use Local or Cloud APIs Based on Client Needs Instead of cloud voice APIs that cost $0.05/request and require sending user audio to third-party servers, I deploy local audio models (74–466MB) running entirely in-browser — works fully offline, and significantly lower operating costs over time. ✅What this means for you: If you're building voice features, there's a path to save substantially on API costs while giving your users genuinely better privacy. ✔️ I Orchestrate Multi-AI Systems — Not Just One API Call I architect intelligent pipelines: for speed, for privacy, for fallbacks — with smart validation and context-aware processing at every layer. This means your system doesn't break when one provider has an outage. ✅What this means for you: A resilient AI architecture that keeps working even when individual providers go down, and gives you flexibility to optimise costs over time. ✔️ I Build Complete Products — Not Just the AI Part Authentication, db architecture, real-time dashboards, payment integrations, deployment pipelines, monitoring — I handle the full stack. You don't need to coordinate between three different freelancers to get a working product. ✅What this means for you: One hire. One person is accountable for the entire system. Less coordination overhead, fewer things falling through the cracks. ✔️ I Stay — 18+ Months of CTO-Level Work Architecture, development, DevOps, security, and strategic decisions. 40% reduction in infrastructure costs—99.9% uptime. ✅What this means for you: I build systems I'm willing to maintain long-term. You're not hiring someone who disappears after delivery — you're bringing on a technical partner. ✅𝗥𝗘𝗖𝗘𝗡𝗧 𝗣𝗥𝗢𝗗𝗨𝗖𝗧𝗜𝗢𝗡 𝗪𝗢𝗥𝗞 ✔️ MedGemma FORENSIC: 4.3GB LLM/Vision/Audio models running offline on 4GB Androids. Submitted to Google's Kaggle Competition. ✔️ serper-search-mcp: Published enterprise MCP server on NPM with Official Google's 5 search types and full Docker support. ✔️ SM Expense: Privacy-First SaaS with Local Voice input & OCR (90% accuracy in 3s). ✅𝗪𝗛𝗔𝗧 𝗬𝗢𝗨 𝗚𝗘𝗧 📱 Edge AI & Voice ✔️ Offline LLM on Android — Kotlin, custom memory kernels, constrained hardware ✔️ P2P mesh networking for zero-connectivity environments ✔️ Native Android AI apps — Kotlin, Jetpack Compose ✔️ RAG systems — Pinecone, ChromaDB, FAISS, PGVector ✔️ AI agents — MCP servers, LangChain, LangGraph ✔️ Multi-modal AI — voice, vision, text orchestrated together ✔️ Privacy-first AI — on-device processing, no data leaks, browser-based, fully offline ✔️ Complete AI SaaS MVPs — concept to production in 4–8 weeks ✔️ Fractional CTO for AI startups ✅𝗙𝗨𝗟𝗟-𝗦𝗧𝗔𝗖𝗞 & 𝗦𝗔𝗔𝗦: ✔️ Complete SaaS applications — Next.js, React, TypeScript, PostgreSQL ✔️ Real-time dashboards and analytics — data visualisation and business insights ✔️ Authentication and multi-tenancy — NextAuth, Prisma, role-based access control ✔️ DevOps and infrastructure — Docker, AWS, CI/CD, monitoring, cost optimisation ✅𝗧𝗘𝗖𝗛 𝗦𝗧𝗔𝗖𝗞 ✔️AI: OpenAI, Claude, Gemini, Groq, LangChain, LangGraph, MCP, HuggingFace ✔️Vector: Pinecone, FAISS, ChromaDB, PGVector ✔️Edge: Kotlin, Android SDK, Jetpack Compose, TFLite, P2P mesh ✔️Frontend: Next.js, React, TypeScript, Tailwind CSS, Laravel ✔️Backend: Node.js, Python (FastAPI),PostgreSQL, Redis, Prisma ✔️Infra: AWS, Docker, GitHub Actions, Sentry ✅𝗣.𝗦. Not sure whether you need full on-device AI or a lean cloud integration? If you're building something technically interesting, I'd genuinely enjoy hearing about it. Send me a quick message describing your core problem — I'll give you an honest technical assessment of what's feasible on your real hardware.

💡Machine Learning Engineer with Expertise in Computer Vision, Deep Learning, Natural Language Processing, Generative AI, AWS cloud services and more!💡 Greetings, I bring over four years of expertise in machine learning and AI, specializing in cutting-edge ML, Deep Learning, and Computer Vision domains. My proficiency spans diverse AI areas, including Computer Vision, Deep Learning, Natural Language Processing, Reinforcement Learning, Generative AI, and LLM. I've strategically applied AI in medicine, academics, sports, and industry, streamlining processes such as data annotation, model training, and deployment. Let's collaborate to elevate your projects with innovative AI solutions. 🌟 Key Services: - Data Cleaning & Preprocessing - Data Scraping & Annotation - Deep Learning Model Training & Optimization - Custom Data and Model Building, Training & Integration - Actionable Recommendations & Insights - Language Vision Models(LLaVa,etc) - GPT-3/4-based agent workflows - Prompt Engineering - API Endpoint Development & Integration 💡 Projects: - Tennis serve Detection - Object Counting and direction detection - Face Mask Detection - Plant Disease Detection - Smart attendance using Face Recognition - Artificial Face emotion generation - Image Tag Detection - Bengali Handwritten Grapheme Classification - Leaf Disease Classification - kidney functional tissue units (FTUs) detection - Hungry Geese agent training 🚀 Expertise: - ML Frameworks: PyTorch, Tensorflow, PyTorchLightning - API Frameworks: FastAPI, Flask - Computer Vision (CNN, Yolo, Yolo-nas, Faster/Fast-RCNNs, ResNet, MobileNet, DeepSort, Semantic & Instance Segmentation, BotSort, Bytetracker, Vision Transformer, etc.) - GANs, Audio Classification, Recommendation & Forecasting Models, Text-to-Speech, Speech-to-Text, Supervised and Unsupervied ML, Reinforcement Learning - BERT, Transformers, GPT-3/4, ChatGPT, Langchain - CLIP, DALL-E, Stable Diffusion, Image-to-Text, Text-to-Image, StyleGAN, CycleGAN - AWS Services (Textract, Rekognition, etc.) - NLP, Time Series Analysis, Machine Learning Algorithm Integration - Amazon SageMaker, Textract, Transcribe, Translate, Rekognition, Comprehend - AWS S3, AWS Lambda, EC2 machine, SNS, SQS, CodeBuild, VPC - GCP compute Engine, GCP VPC - PostgreSQL database, AWS dynamodb, MySQL 🛠 Technical Skills: - Languages: Python, C++, Reactjs - Libraries: OpenCV, Numpy, Scikit-Learn, Matplotlib, SciPy, Pandas, Spacy, NLTK, XGBoost - Version Control: GIT, Github, Bitbucket - Analysis: Statistics, Calculus, Classification/Clustering, Machine Learning, Deep Learning - Methodologies: Scrum, Kanban, TDD - CI/CD: Github Action, Jenkins, Kubernetes - Other: Linux, Strong Problem Solving Background Thank you for considering my profile. I'm passionate about the endless possibilities of machine learning to solve real-world challenges. If you're ready to bring your AI ideas to life, let's connect. I'm eager to discuss your project and how my expertise can contribute to your success. Don't hesitate to reach out—I'm here to help you innovate and make your vision a reality🚀.

Data science and machine learning professional with more than 7 years of experience. Currently, building and evaluating large language model based solutions for healthcare and other industries. Previously, I was the Lead Data Scientist, at a startup. During that time, I have gathered experience in the Power and Energy, Fashion/Apparel, FMCG, Manufacturing, Retail, Open Government Data, Banking and Finance industries. I started working as a Freelance Data Scientist in 2021, and have worked for a healthcare company for 1 year, where I have led the efforts to automate complex data pipelines for health, growth and performance metrics. My responsibilities in my previous jobs included leading and mentoring data science teams, planning projects and project sprints, developing statistical and machine learning models and pipelines, data visualization/dashboards, data cleaning automation scripts. I engineered and automated a data pipeline for operational and financial dashboards for Bangladesh's public power sector. The project was awarded Bangladesh National ICT award in Big Data Category in October 2019. I also worked on Natural Language Processing problems for a flagship product, where natural language questions are translated to valid SQL queries. More recently, I spearheaded efforts to win clients in the banking sector by contributing in the development of demo applications like customer segmentation model, credit scoring model, bank statement OCR, and management dashboards. I orchestrated collecting national Covid-19 data from various sources and implemented a database and a data warehouse on the Google Cloud Platform. Local policymakers and other researchers consumed these data for analysis and decision-making. Then, at a California-based unicorn startup, I performed offline evaluations of production machine learning models to analyze search and ranking performance. I also tested and proposed new features for ranking models that match millions of developers on our platform to thousands of jobs. I am passionate about my work and love exploring new problems, which is why I enjoy working as a Data Science Freelancer. 🚀Skilled in: - Python - PyTorch - HuggingFace - OpenAI APIs - LangChain - TensorFlow - Keras - PySpark - OpenCV - FastAPI - Flask - Streamlit - Git - Google Cloud Platform - Amazon Web Services - SQL - Microsoft Power BI - Google Data Studio - Mode Analytics - Apache Superset - Jupyter - Plotly - Dash - DAX

I am Saman Sarker Joy, an AI/ML Engineer with experience in Deep Learning, NLP and Computer Vision. My expertise includes fine-tuning Large Language Models (LLMs), OCR Technologies, Image Processing and AI-powered automation. I have worked as a Machine Learning Engineer, contributing to AI-driven OMR systems and academic search engines and as a Research Assistant in Machine Learning & Deep Learning projects. 🔹 My Expertise & Key Projects: ✅ Large Language Models (LLMs) & NLP - Fine-tuned Llama, Gemma, BERT, Electra Models for NLP applications. - Developed Bangla Clickbait Detection & Fake News Detection using domain-adaptive BERT. - Created Bangla Text-to-IPA transcription using Seq2Seq models for phonetic accuracy. - Developed Gazetteer-Enhanced Bangla NER using BERT & CRF for improved semantic entity extraction. - Worked with low-resource NLP models, integrating unsupervised clustering techniques. ✅ Computer Vision & Image Processing - Built synthetic cell segmentation models using Custom Deep Learning models, OpenCV and YOLO Models. - Developed a Grapheme Detection System for Bangla handwriting recognition using VGG-19. Worked on OMR Scanning for AI-powered exam platforms, increasing accuracy to 96%. ✅ AI-Driven Automation & Backend Development - Led Testiphy, an AI-powered OMR & Gamified Exam Platform at RubizCode. - Built Taalaash, an academic search engine using OCR, NLP and AI-driven search optimization. - Developed a Question Bulk Upload system using Next.js, Flask, OpenAI APIs, and Beautiful Soup, increasing efficiency 50x. 🔹 Technical Skills: ✅ Programming & Libraries: Python, C++, JavaScript, TypeScript, Flask, Next.js ✅ NLP & LLMs: BERT, Electra, GPT, NLTK, SpaCy, Hugging Face, OpenAI APIs ✅ Data Science: Pandas, NumPy, Matplotlib, Seaborn ✅ Computer Vision: OpenCV, Image Segmentation, OCR, Object Detection, YOLO ✅ ML & DL Architectures: CNN, RNN, LSTMs, Transformers, Attention Mechanisms ✅ DevOps & Cloud: Docker, Microsoft Azure, Google Cloud 🔹 Why Work With Me? ✔️ Industry & Research Expertise – Experience in AI product development & academic research. ✔️ Proven Impact – Built AI-driven automation solutions, scaling efficiency 50x. ✔️ Strong Research Background – Published peer-reviewed NLP & CV papers. ✔️ Communication & Collaboration – Ensures clear updates, well-documented code, and transparent workflows. If you need expertise in Deep Learning, NLP, LLMs or Computer Vision, let's collaborate! 💬 Let’s discuss your project today!