AI Voice Agent Developer
AI Voice Agent Developer<p><strong>Company:</strong> OneNDF</p><p><strong>Location:</strong> New Delhi</p><p><strong>Experience:</strong> 0–2 Years | Strong Freshers Welcome</p><p><strong>Employment Type:</strong> Full-time Not a work from home</p>About OneNDF<p>OneNDF is a fintech platform focused on <strong>Home Loans and Loan Against Property</strong>, combining lending intelligence, lender matchmaking and AI-led customer engagement.</p><p>We are building AI agents that can interact with customers, partners and bank/NBFC representatives through <strong>voice and WhatsApp</strong>, understand natural conversations, collect information intelligently and take the conversation forward based on context.</p>About the Role<p>We are looking for an <strong>AI Voice Agent Developer</strong> who is excited about building conversational AI systems that can talk to real customers.</p><p>This is a hands-on development role.</p><p>You will work on AI voice agents that can:</p><ul><li>Conduct natural two-way conversations</li><li>Understand English, Hindi and Hinglish</li><li>Ask intelligent follow-up questions</li><li>Handle interruptions and changing conversation paths</li><li>Extract structured information from calls</li><li>Use business data and APIs during conversations</li><li>Trigger workflows after a call</li><li>Improve continuously based on actual conversations</li></ul><p>The objective is not to build a simple IVR or scripted bot. We want to build <strong>AI agents that can understand the context of a conversation and respond naturally.</strong></p>What You Will Work OnReal-Time AI Voice Agents<ul><li>Build and improve real-time conversational voice agents.</li><li>Integrate Speech-to-Text, LLM and Text-to-Speech systems.</li><li>Build low-latency voice pipelines.</li><li>Handle interruptions, silence, turn-taking and natural conversation flow.</li><li>Develop Hindi, English and Hinglish conversational capabilities.</li><li>Improve voice-agent latency and response quality.</li></ul>Conversational Intelligence<ul><li>Design conversation flows and AI-agent behaviour.</li><li>Build prompt and context-management systems.</li><li>Enable agents to remember information collected earlier in a conversation.</li><li>Make agents ask relevant follow-up questions rather than blindly following scripts.</li><li>Build objection-handling and conversation-branching logic.</li><li>Identify customer intent and dynamically change the conversation.</li></ul>LLM & Tool Integration<p>Integrate voice agents with:</p><ul><li>LLM APIs such as OpenAI and Gemini</li><li>CRM systems</li><li>Customer databases</li><li>Lending calculators</li><li>Internal APIs</li><li>WhatsApp workflows</li><li>Knowledge bases</li><li>Lead-management systems</li></ul><p>Enable the AI agent to call tools during conversations—for example, checking an eligibility calculation or retrieving relevant information before responding.</p>Call Intelligence<p>Build systems to automatically capture:</p><ul><li>Call transcripts</li><li>Customer requirements</li><li>Key information collected</li><li>Customer objections</li><li>Follow-up actions</li><li>Conversation outcome</li><li>Lead qualification</li><li>Structured CRM fields</li></ul><p>Create mechanisms to convert an unstructured conversation into usable structured data.</p>Freshers Can Apply<p>We are open to strong freshers who have built projects involving:</p><ul><li>AI agents</li><li>Voice bots</li><li>Chatbots</li><li>LLM applications</li><li>Python automation</li><li>OpenAI / Gemini APIs</li><li>Real-time applications</li></ul><p>A working project is more valuable to us than only having certificates.</p><p>If you have built something using AI, <strong>show it to us.</strong></p>What Will Make You Stand Out<p>You will stand out if you have:</p><ul><li>Built an AI voice bot yourself</li><li>Connected an LLM with an external API</li><li>Worked with streaming audio</li><li>Built function/tool calling with LLMs</li><li>Experimented with RAG or conversational memory</li><li>Deployed an AI application</li><li>Used GitHub to showcase your projects</li><li>Tried solving latency or hallucination problems in LLM applications</li></ul>The Kind of Problems You Will Solve<p>Examples of problems you may work on:</p><p><strong>How does an AI agent know when the customer has finished speaking?</strong></p><p><strong>How should the agent react when the customer interrupts it?</strong></p><p><strong>How can the agent understand a sentence that mixes Hindi and English?</strong></p><p><strong>How can the agent collect ten pieces of information naturally without sounding like a questionnaire?</strong></p><p><strong>How can the AI call a calculator or CRM API during the conversation and use the result in its response?</strong></p><p><strong>How do we prevent the agent from giving information it does not know?</strong></p><p><strong>How do we analyse hundreds of calls and automatically improve the conversation flow?</strong></p><p>These are the kinds of real-world conversational AI problems you will work on at OneNDF.</p>What You Will Learn<p>You will get hands-on exposure to:</p><ul><li>Generative AI</li><li>AI Agents</li><li>Real-time Voice AI</li><li>LLM orchestration</li><li>Prompt engineering</li><li>Tool / Function Calling</li><li>Speech AI</li><li>Conversational Design</li><li>API integrations</li><li>Fintech</li><li>CRM automation</li><li>Production deployment of AI applications</li></ul>Who Should Apply<p>This role is ideal for someone who enjoys <strong>building and experimenting</strong>.</p><p>You don't need years of experience.</p><p>We are looking for someone who sees a new AI technology and immediately thinks:</p><p><strong>“Let me build something with this and see if it works.”</strong></p><p>If you are curious about where <strong>AI Agents + Voice + Real-world business workflows</strong> are heading, this role gives you the opportunity to build them in production.</p>