Speech Recognition Jobs

Train OpenWakeWord with a Custom Wake Word (High Accuracy, Low False Positives) Project Overview I am looking for an experienced AI/ML engineer to train OpenWakeWord with a custom wake word. The goal is to build a production-quality wake word model that is highly accurate, responsive, and resistant to false activations. Scope of Work * Train a custom OpenWakeWord model from scratch or fine-tune an existing pipeline. * Use real positive voice samples from multiple speakers, accents, speaking speeds, distances, and environments. * Collect and use a large negative dataset containing similar-sounding words, random conversations, TV/audio, background noise, music, and other speech to minimize false positives. * Optimize the model for: * High detection accuracy * Extremely low false ...

## **Native Malaysian English Speakers Needed – AI Voice Recording Project ($12/hour)** **TRS Multilingual** is looking for contributors with **fluent Malaysian English (Malay-accented English)** to join our AI voice data collection project. The task is simple: record natural, spontaneous speech in English with your natural Malaysian accent and upload the recordings through our platform. ### **About the Project** This project helps improve AI speech recognition and transcription systems. Participants will record natural spoken English using a **Malaysian accent**, covering everyday topics such as personal experiences, daily life, hobbies, travel, food, technology, and casual conversations. ### **Responsibilities** * Record natural spoken English with a **Malaysian accent** (10&...

I need a lightweight, secure application that can listen to both sides of my life-insurance sales calls, transcribe them in real time, and drop only the relevant facts into a structured worksheet on-screen. While the caller is still speaking, the system should automatically recognise and populate fields such as full name, date of birth, address, phone number, current medications and medical conditions. Key behaviour • Ignore small talk and fillers; capture only what belongs on the application. • If the caller corrects a detail (“actually, my phone is 555-0199”), the field must update instantly. • Anything that sounds uncertain—uncommon drug names, hard-to-spell surnames, email addresses, partial street names—should be highlighted or flagged s...

I need ready-made, professionally recorded conversational speech that I can license immediately for AI training. All material must be dual channel (each speaker isolated), feature spontaneous human dialogue, and come with accurate time-aligned transcriptions plus basic speaker metadata. Languages and expected hours per language are as follows: • Spanish – US (es-US): 220h • French – Canada (fr-CA): 220h • Arabic – UAE (ar-AE): 220h • Italian (it-IT): 220h • Thai (th-TH): 195h • Turkish (tr-TR): 220h • Hebrew (he-IL): 220h • Dutch (nl-NL): 220h • Greek (el-GR): 220h • Vietnamese (vi-VN): 220h The recordings should cover a healthy mix of customer service interactions, casual conversations, and technic...

I have a series of recorded, real-life conversations in Tamil that need to be transformed into accurate, readable text. Each audio clip involves at least two speakers, so I need every word captured verbatim and a clear timestamp inserted whenever the speaker changes. Please deliver a separate UTF-8 .txt file for each recording. Keep the layout simple: speaker label, timestamp, then the spoken line. Consistent spelling, tidy punctuation, and faithful reproduction of colloquial phrasing are all important to me. The recordings were made across various everyday settings in India, so you will occasionally encounter background noise or regional inflections; an attentive ear for Tamil dialects is essential. I value precision over speed, but a 24–48 hour turnaround per 60-minute file is my...

Transcript of audio *no enhancement* Audio transcribed as is! Supposedly this audio has been enhanced and should be pretty clear I have several short (less than 1minute) clips that I need to know what is being said. I have a few of the files attached so that you can hear a sample. Thank you for your time

we want you to built a phone assistance system for voice and chat and there are many features to build. • Answering questions in real time • Setting reminders and alarms with contextual follow-ups • Making calls and sending messages hands-free • Language translation on demand and many many other features Beyond those must-haves, I’m open to additional smart features you can suggest and implement as long as they blend seamlessly into the same voice interface. The ideal workflow: 1. Architect the solution (cloud or on-device), outlining key tech such as Python, TensorFlow, Dialogflow, Twilio, or any stack you believe is best for low-latency, privacy-minded processing. 2. Deliver a functional demo—APK, iOS TestFlight build, or hardware prototype...

You’ll be recording a set of short, everyday Tamil phrases that will become part of a larger voice-recognition training corpus. I will supply the full phrase list; each line is just a few words, so a normal speaking pace should keep the entire session brief. What I need from you • A native (or near-native) Tamil accent from India. • A quiet space and a smartphone or USB microphone capable of producing clear, hiss-free audio. • One clean, separate audio file per phrase—raw WAV or high-quality MP3 is fine. • Correct pacing: natural, conversational speed with no exaggerated pauses. How the task works I’ll send over the script and simple recording guidelines. You read each everyday conversation phrase once, press stop, and move to the next. When yo...

SPPARKAL – AI Sewing Machine Maintenance Assistant Develop a voice-based mobile app called SPPARKAL for sewing machine maintenance and troubleshooting. The user should speak naturally and describe the machine problem. The app should convert speech to text, identify the language and problem, match it with a pre-fed and approved SPPARKAL troubleshooting database, and provide step-by-step guidance in the same language. Initially support Marathi, Hindi and English. The AI should only understand, classify and match the user's problem; it must not invent technical solutions. All troubleshooting procedures will be pre-fed and structured with problem name, symptoms, safety instructions, steps, user responses, next steps and escalation. The app should maintain conversation context and...

I’m coordinating a speech-data study that trains medical-domain AI models, and I need your help sourcing authentic voices from the healthcare community. The task itself is straightforward: each recruited participant—licensed or currently employed clinicians, nurses, therapists, medical technologists, or similar—records a short series of question-and-answer prompts in their native environment. Because we are focusing on question and answer sessions, every clip must clearly capture both the spoken question and the spontaneous response. Here’s what I’ll look for from you: • Identify and confirm genuine healthcare professionals willing to contribute. • Guide them through my simple recording script (I supply the Q&A prompt list and a quick onboa...