Senior AI/ML Engineer – Conversational AI & Voice Platform (BFSI)
Company: LeadFlowNex AI
Location: Remote / New Delhi
Job Type: Full-time
Experience: 5+ years
About LeadFlowNex AI
LeadFlowNex AI is building an advanced Conversational AI Voice Platform for BFSI, designed to automate high-volume outbound customer conversations for banking, lending, insurance, and financial services.
Our platform enables AI voice agents to communicate naturally with customers, understand conversations, collect information, qualify leads, perform eligibility assessment, and trigger business workflows in real time.
We already have a complete software development team covering backend, frontend, integrations, and product engineering.
We are now looking for a Senior AI/ML Engineer to own the AI/ML layer of the platform.
This is a core technical role. You will be responsible for building, training, fine-tuning, evaluating, and deploying the AI models that power our conversational voice platform.
What You Will Own1. AI/ML Model Development
- Design and develop custom AI/ML models for our platform.
- Train, fine-tune, evaluate, and optimize machine-learning and deep-learning models.
- Work with our existing datasets to build production-ready models.
- Perform data preprocessing, feature engineering, labeling validation, and dataset optimization.
- Select appropriate architectures and approaches based on business and technical requirements.
- Build reproducible training and evaluation pipelines.
2. Conversational AI & NLP
- Develop conversational intelligence for AI voice agents.
- Build intent classification, entity extraction, semantic understanding, and conversation-state models.
- Improve understanding of Hindi, English, Hinglish, and other Indian languages.
- Work with LLMs and open-source models for conversational AI.
- Implement prompt engineering, structured outputs, RAG, tool/function calling, and fine-tuning where appropriate.
- Develop methods to reduce hallucination and improve factual consistency.
3. Voice AI
- Work with real-time Speech-to-Text (STT), Text-to-Speech (TTS), VAD, turn detection, and streaming speech pipelines.
- Integrate and optimize technologies such as Whisper, Deepgram, Google Speech-to-Text, Azure Speech, ElevenLabs, or equivalent platforms.
- Improve latency, accuracy, interruption handling, and naturalness of AI conversations.
- Work closely with our existing engineering team on telephony and real-time communication integration.
4. AI Decisioning & Predictive Models
- Build models for eligibility, lead qualification, propensity, scoring, ranking, and routing.
- Develop predictive models using our existing BFSI datasets.
- Combine machine-learning predictions with deterministic business rules.
- Build real-time decision-support systems for loan, insurance, and credit-card workflows.
- Monitor model performance and continuously improve prediction quality.
5. Model Evaluation
Design and maintain rigorous evaluation frameworks for:
- Model accuracy
- Precision, recall, F1-score
- Intent recognition
- Entity extraction
- STT accuracy / Word Error Rate
- Conversation success rate
- Response latency
- Hallucination/error rate
- Lead qualification rate
- Business conversion metrics
- Model drift
The objective is not just to build models that perform well in experiments, but to build AI that performs reliably in real customer conversations at production scale.
6. Production AI & MLOps
You will work with our existing development and DevOps team to:
- Deploy trained models into production.
- Build model inference APIs and services.
- Optimize inference latency and infrastructure cost.
- Implement model versioning and monitoring.
- Establish retraining and continuous-improvement pipelines.
- Support high-concurrency real-time AI workloads.
- Work with AWS/cloud infrastructure, Docker, CI/CD, and monitoring systems.
What We Already Have
You will not be joining as a one-person development team.
We already have:
- Python Full-Stack Developers
- Backend development
- Frontend development
- Product engineering
- API/integration development
- Database infrastructure
- Development support
- Existing project datasets
- Product/project requirements
Your primary responsibility will be the AI/ML layer.
You will work closely with the existing engineering team to integrate your models into the production platform.
What We're Looking ForRequired
- 5+ years of professional experience in AI/ML Engineering, Machine Learning, Data Science, or a closely related field.
- Strong hands-on experience with Python.
- Proven experience training and deploying ML/DL models.
- Experience with PyTorch, TensorFlow, Scikit-learn, or equivalent frameworks.
- Strong understanding of NLP and modern AI architectures.
- Practical experience with LLMs and Generative AI.
- Experience with model fine-tuning and/or transfer learning.
- Strong understanding of model evaluation and experimentation.
- Experience taking models from development to production.
Mandatory Model-Building Experience
This is not an AI API integration role.
Candidates must have hands-on experience with:
Data → Training → Fine-tuning → Evaluation → Optimization → Deployment
Candidates whose experience is primarily limited to integrating OpenAI, Gemini, or other third-party AI APIs without hands-on model development/training experience will not be suitable for this position.
Strongly Preferred
- Conversational AI experience
- Voice AI experience
- STT/TTS experience
- LLM fine-tuning
- LoRA / QLoRA
- Hugging Face Transformers
- PyTorch
- RAG
- Agentic AI
- Hindi/Hinglish NLP
- Real-time AI systems
- BFSI/FinTech experience
- Predictive/risk/propensity modeling
- AWS/GCP/Azure
- Docker/Kubernetes
- MLOps
Nice to Have
- Experience with Whisper, Deepgram, ElevenLabs, Azure Speech, or similar technologies.
- Experience with Twilio, SIP, WebRTC, VoIP, or telephony systems.
- Experience processing large-scale call transcripts or voice datasets.
- Experience building AI agents for sales, customer support, collections, lending, or lead qualification.
- Experience with LangChain, LlamaIndex, CrewAI, or similar frameworks.
- Experience with Indian-language conversational AI.
Education
Bachelor's or Master's degree in:
- Computer Science
- Artificial Intelligence
- Machine Learning
- Data Science
- Engineering
- Mathematics
- Statistics
or a related technical field.
Strong practical experience can outweigh academic qualifications.
First 90 Days
The selected candidate will be expected to make significant progress toward a production-ready AI system within the first three months.
Month 1
- Audit existing datasets
- Define AI/ML architecture
- Establish baselines
- Prepare training pipelines
- Develop initial models
- Establish evaluation metrics
Month 2
- Train/fine-tune production models
- Improve conversational intelligence
- Develop decisioning/predictive models
- Integrate models with the existing engineering stack
- Conduct extensive model evaluation
Month 3
- Production deployment
- Performance and latency optimization
- Load/concurrency testing
- Monitoring and model versioning
- Final AI/ML integration
- Production pilot readiness
What We Offer
- Opportunity to build a real-world Conversational AI + Voice AI platform from the ground up.
- Significant ownership of the AI/ML architecture.
- Existing engineering team to support implementation and integration.
- Existing datasets and product infrastructure.
- Direct collaboration with the Founder and technical team.
- High autonomy and technical ownership.
- Competitive compensation.
- ESOP/equity opportunities for the right candidate.
- Opportunity to work on AI applications in the BFSI/FinTech sector.
How to Apply
Please submit:
- Updated CV/Resume
- GitHub/portfolio, if available
- Brief description of the most complex ML/AI model you have personally built and trained
- Details of any production AI/ML system you have deployed
- Examples of LLM, NLP, or Voice AI work, if applicable
LeadFlowNex AI is an equal opportunity employer.
Pay: ₹60,000.00 - ₹90,000.00 per month
Work Location: Hybrid remote in Indian Habitat Centre, Delhi
