AI Developer

Experience: 4-8 years

Location: MUMBAI

Domain: BFSI

Please share your resume to [Confidential Information]

PwC India is seeking an experienced AI Developer with a strong background in the BFSI (Payments) domain to design, build, and implement AI/ML-based solutions across our applications, with a focus on Payments, Fraud Detection, Compliance and Generative AI (GenAI) applications. The ideal candidate will have proficiency in developing AI-driven applications, fine-tuning LLMs, building Retrieval-Augmented Generation (RAG) chatbots, and integrating AI and GenAI within payment systems. Hands-on experience with Llama, Llamaindex, Langchain, Huggingface, and vector databases is essential.

This role involves working closely with client teams to deploy cutting-edge AI/ML technologies, driving innovation, efficiency, and security in the BFSI space.

---

Key Responsibilities:

1. AI and ML Integration:

o Develop and integrate AI and ML models into BFSI applications, focusing on Payments and Fraud Detection.

o Implement algorithms for fraud prevention, risk scoring, and real-time transaction monitoring.

o Design predictive models for customer behavior, fraud patterns, and payment trends.

2. GenAI Application Development:

o Build Generative AI (GenAI) applications for personalized customer experiences and support within the BFSI ecosystem.

o Fine-tune LLMs such as Llama and Huggingface models for domain-specific tasks (e.g., automated customer service, financial advisory, risk analysis).

o Utilize Langchain to develop context-aware, intelligent chatbots and virtual assistants.

3. RAG-based Chatbot Development:

o Design and implement Retrieval-Augmented Generation (RAG) chatbots to enhance real-time, accurate information retrieval for customer and internal queries.

o Leverage Llamaindex and Langchain for efficient information retrieval and real-time contextual responses, particularly in payment processes and fraud management systems.

4. AI in Payments, Fraud Management, Risk & compliance managment:

o Collaborate with partner / client teams to integrate AI and GenAI models for optimizing payment processing and reducing technical declines.

o Implement AI-based fraud detection algorithms to detect and block suspicious transactions while minimizing false positives.

5. LLM Fine-tuning and Customization:

o Fine-tune LLMs such as Llama, customizing models for BFSI-specific tasks, ensuring precision, performance, and scalability.

o Work with Llamaindex to index and retrieve relevant financial data, and integrate LLMs for real-time insights.

6. Vector Database and AI Integration:

o Utilize vector databases (e.g., Pinecone, Weaviate) for managing large-scale, high-dimensional data needed for embedding and storing vectors from AI models.

o Leverage vector search capabilities to improve real-time information retrieval and similarity searches within fraud detection and payment systems.

7. Collaborative Innovation:

o Collaborate to create propositions on AI / GenAI for clients.

o Ensure AI implementations adhere to regulatory standards.

8. Research and Development:

o Stay updated on the latest trends in AI, ML, and GenAI technologies and their BFSI applications.

o Pilot emerging AI techniques, such as explainable AI (XAI), to improve transparency in fraud detection models.

Required Skills and Qualifications:

4+ years of experience in AI/ML development within the BFSI domain.

Hands-on experience with LLMs such as Llama, Huggingface, and frameworks like Langchain and Llamaindex.

Experience with vector databases like Pinecone, Weaviate, or similar.

Strong proficiency in machine learning frameworks (e.g., TensorFlow, PyTorch, Scikit-learn).

Proven expertise in developing AI/ML solutions for Payments and Fraud Detection systems.

Programming skills in Python or Scala.

Familiarity with cloud platforms (AWS, Azure, GCP) for AI/ML solution deployment.

Solid understanding of NLP, LLMs, Generative AI, and RAG techniques.

Knowledge of BFSI regulations.