Agentic AI Project Lead Voice AI Solutions BFSI

Location: Gurgaon, India
Experience: 610 years overall, with at least 3 years in Conversational AI, Voice AI, or Agentic AI
Employment Type: Full-time
Work Mode: Work from Office

About the Role

We are looking for an experienced Agentic AI Project Lead to lead the design, development, and deployment of AI-powered voice solutions for the Banking, Financial Services, and Insurance (BFSI) industry.

The ideal candidate must have hands-on experience building and deploying AI voice agents capable of conducting natural, human-like conversations with customers, understanding their intent, taking actions through backend systems, and completing business workflows autonomously within defined guardrails.

The role will focus primarily on digital lending, personal loans, loan collections, customer onboarding, loan application follow-ups, insurance sales, policy renewals, and customer support automation.

This is a hands-on technical leadership role requiring a combination of Agentic AI expertise, voice technology, software engineering, BFSI domain knowledge, and project delivery experience.

Key Responsibilities

1. Agentic AI and Voice Solution Development

• Lead the architecture, development, and deployment of production-grade AI voice agents for BFSI use cases.
• Build agentic workflows that can understand customer intent, maintain conversational context, make tool calls, execute business actions, and manage multi-step interactions.
• Develop inbound and outbound AI calling solutions for loan origination, lead qualification, customer verification, EMI reminders, collections, insurance renewals, and customer servicing.
• Implement multi-agent architectures where appropriate, including orchestration, task delegation, escalation, and workflow execution.
• Design robust conversation flows with interruption handling, contextual memory, multilingual support, retries, fallback mechanisms, and human-agent handoff.
• Optimize voice agents for low latency, natural speech, accuracy, reliability, and customer experience.

2. Voice AI Technology and Integrations

• Work with speech-to-text (STT), text-to-speech (TTS), large language models (LLMs), telephony platforms, and real-time voice streaming technologies.
• Integrate AI voice agents with telephony providers, SIP, cloud calling infrastructure, contact-centre platforms, and CRM systems.
• Implement real-time voice pipelines involving audio streaming, speech recognition, response generation, and speech synthesis.
• Build integrations with loan origination systems (LOS), loan management systems (LMS), CRM platforms, payment gateways, insurance systems, and customer databases.
• Evaluate and integrate suitable voice and AI platforms, such as LiveKit, Vapi, Retell AI, Twilio, Deepgram, ElevenLabs, and comparable technologies, based on project requirements.
• Develop APIs and backend services that enable voice agents to retrieve customer information, update records, schedule callbacks, initiate approved workflows, and log call outcomes.

3. BFSI Domain Solutions

Lead the development of AI solutions for the following business use cases:

Digital Lending and Personal Loans

• Lead qualification and customer profiling.
• Personal loan eligibility pre-screening and application assistance.
• Document collection reminders and application follow-ups.
• KYC process guidance and status updates.
• Loan application abandonment recovery.
• EMI reminders, repayment follow-ups, and payment assistance.
• Customer queries related to loan status, repayment schedules, and outstanding balances.

Collections and Recoveries

• Automated outbound calling for early-stage delinquency and payment reminders.
• Promise-to-pay capture and follow-up scheduling.
• Payment link sharing through approved channels.
• Structured conversations based on customer responses and repayment status.
• Escalation of disputes, hardship cases, complaints, and sensitive interactions to human agents.
• Monitoring of call quality, adherence to approved scripts, and collections policies.

Insurance

• Lead qualification for insurance products.
• Policy renewal reminders and follow-ups.
• Customer onboarding and policy servicing.
• Premium payment reminders and assistance.
• Claims status updates and documentation follow-ups.
• Customer queries relating to policy coverage, exclusions, and renewal requirements, with appropriate escalation where necessary.

4. Technical Architecture and Engineering

• Design scalable, secure, and resilient AI application architectures using Python, FastAPI, Node.js, REST APIs, and event-driven services as appropriate.
• Implement LLM integrations, function calling, structured outputs, retrieval-augmented generation (RAG), and workflow orchestration.
• Work with agent frameworks and orchestration technologies such as LangGraph, LangChain, Semantic Kernel, or equivalent solutions.
• Develop integrations with databases, vector stores, cloud infrastructure, and enterprise systems.
• Implement session management, conversation state, audit logs, retries, observability, and failure recovery.
• Design evaluation frameworks to test response accuracy, tool execution, latency, hallucinations, task completion, and overall call quality.

5. Project Leadership and Delivery

• Own the complete project lifecycle, from requirement gathering and solution design to development, testing, deployment, and production support.
• Translate business requirements into technical architecture, implementation plans, milestones, and measurable acceptance criteria.
• Lead developers, AI engineers, backend engineers, and QA teams.
• Coordinate with product managers, BFSI stakeholders, compliance teams, and operations teams.
• Manage delivery schedules, dependencies, risks, production issues, and stakeholder communication.
• Conduct technical design reviews, code reviews, sprint planning, and release readiness assessments.
• Drive continuous improvement based on call analytics, customer feedback, and operational metrics.

6. Security, Governance, and Compliance

• Build solutions with appropriate data privacy, access controls, encryption, and auditability.
• Implement customer consent management, approved calling practices, and applicable Do Not Disturb / opt-out requirements.
• Ensure that AI agents follow approved scripts, business rules, and escalation policies.
• Design safeguards against prompt injection, unauthorized tool execution, sensitive data exposure, and incorrect financial commitments.
• Ensure that loan approvals, insurance underwriting decisions, payment transactions, and other regulated actions remain within approved authorization workflows.
• Work with compliance teams to address applicable RBI, IRDAI, TRAI, and data-protection requirements, depending on the deployment and business model.

Required Technical Skills

• Strong hands-on experience in Agentic AI, Conversational AI, Voice AI, or LLM-powered application development.
• Proven experience building and deploying AI voice agents in production.
• Strong understanding of LLMs, prompt engineering, function calling, tool use, structured outputs, and multi-step agent workflows.
• Experience with real-time voice pipelines, STT, TTS, voice activity detection, streaming, and conversational latency optimization.
• Proficiency in Python and experience building backend services and APIs.
• Experience integrating telephony infrastructure, CRM systems, and enterprise applications.
• Knowledge of agent orchestration frameworks such as LangGraph, LangChain, or equivalent.
• Experience with cloud deployment, containers, monitoring, logging, and production troubleshooting.
• Ability to design automated testing and evaluation for AI-driven conversations.

Preferred Skills

• Experience with voice AI platforms such as LiveKit, Vapi, Retell AI, Twilio, Deepgram, or ElevenLabs.
• Experience with AWS, Azure, or Google Cloud.
• Experience with vector databases, RAG pipelines, and knowledge-base integrations.
• Familiarity with Redis, PostgreSQL, queues, WebSockets, and event-driven architectures.
• Experience with multilingual voice agents, including Indian languages and English.
• Knowledge of call-centre analytics, conversation intelligence, and quality assurance systems.

BFSI Experience Requirements

Candidates must have experience building solutions for at least one of the following:

• Digital lending and personal loan origination.
• NBFCs, fintechs, banks, or digital lending platforms.
• Loan collections, repayment reminders, or debt recovery workflows.
• Insurance sales, renewals, servicing, or claims.
• Contact-centre automation for financial services.

Strong preference: Candidates who have deployed outbound AI calling solutions for personal loan lead conversion, application follow-ups, EMI reminders, or collections.

Key Performance Indicators (KPIs)

• Successful deployment of production-ready AI voice solutions.
• Improvement in call connection, conversation completion, and task completion rates.
• Reduction in average handling time and manual calling effort.
• Accuracy of customer intent recognition and backend tool execution.
• Reduction in hallucinations, incorrect responses, and failed workflows.
• Voice response latency and system availability.
• Compliance adherence, escalation accuracy, and customer complaint rates.
• Project delivery against timelines, budgets, and agreed acceptance criteria.

Candidate Profile

• 6–10 years of overall software engineering, AI, or technology experience.
• At least 3 years of relevant Conversational AI, Voice AI, or Agentic AI experience.
• Demonstrable experience taking at least one AI voice solution from prototype to production.
• Experience leading a technical team or owning end-to-end project delivery.
• Strong problem-solving, communication, stakeholder management, and execution skills.
• Ability to balance rapid AI innovation with production reliability, security, and BFSI compliance.

Interview Evaluation

Candidates should be able to demonstrate or explain:

1. A production AI voice agent they personally built or led.
2. The architecture of the voice pipeline, including STT, LLM, TTS, and telephony.
3. How the agent executes backend actions through APIs and handles failures.
4. How they measure latency, accuracy, hallucinations, and successful task completion.
5. How they design compliant AI calling workflows for lending or insurance.
6. Their specific contribution to development, architecture, deployment, and measurable business outcomes.

Important: Candidates with only chatbot development, prompt engineering, or generic GenAI experience, but no hands-on voice AI deployment experience, will not be considered a strong match for this role.