19 Nov
|
Skeps
|
Gurugram
What You'll Do
- Define the AI/ML architecture. Own the end-to-end design of data ingestion, feature engineering, model training, serving and monitoring. Make sure the platform is scalable, maintainable, secure and able to evolve without rewrites.
- Build ML for lending and payments. Architect solutions for credit risk scoring, fraud and anomaly detection, loan eligibility and offer matching across multiple lenders, and personalisation for merchants and customers.
- Design LLM and GenAI capabilities. Lead RAG pipelines, AI agents, document intelligence and conversational assistants, with strong guardrails for accuracy, privacy and responsible use.
- Own MLOps and model lifecycle. Set up CI/CD for models, experiment tracking, a model registry, automated retraining, drift detection and observability, so models stay reliable after release.
- Integrate with the core platform. Expose AI/ML services to the Node.js and Python microservices through well-designed REST APIs and event-driven pipelines, with idempotency and graceful failure handling for transactional workloads.
- Own security, risk and compliance. Be accountable for data protection, access controls, auditability, explainability and fairness of models that make financial decisions. Partner with legal and compliance on regulatory expectations for digital lending and certifications such as ISO 27001, SOC 2 and PCI-DSS.
- Partner across teams. Work with engineering managers, tech leads, data scientists, product managers, UX designers and business stakeholders to define priorities and tie AI work to measurable business outcomes.
- Write specifications and guide delivery. Develop technical specifications,
review designs and key pull requests, and ensure teams follow agreed architecture and design patterns.
- Drive cost and performance. Manage the cost of training, inference and data infrastructure (FinOps), define SLOs for model-serving latency and accuracy, and lead incident response and post-mortems for AI systems.
- Champion an AI-first approach. Track emerging AI technologies, recommend what to adopt, and help engineers use AI-assisted development in their daily workflows.
- Mentor and grow people. Coach junior engineers and data scientists, raise the technical bar through reviews and pairing, and help build the next layer of technical leaders.
What You Bring
- 10–12 years of software engineering experience, including 4+ years designing and deploying AI/ML solutions in production, with a strong focus on architecture and design.
- Built products from scratch. You have taken end-to-end AI-enabled products to production in a startup or product organisation and lived with the consequences.
- Robust ML foundations. Python with scikit-learn, PyTorch or TensorFlow, XGBoost/LightGBM, Pandas and NumPy. Solid grounding in model evaluation, bias and fairness, and explainability.
- LLMs and GenAI. Prompt engineering, fine-tuning, RAG, embeddings and vector databases,
with frameworks such as LangChain or LlamaIndex.
- MLOps.
Experience with MLflow, Kubeflow, Airflow, SageMaker or equivalent, including model versioning, monitoring and automated pipelines.
- Data and distributed systems. Data pipelines at scale using SQL and NoSQL stores, Spark and Kafka or similar, event-driven design, and idempotency and failure handling for transactional systems.
- Server-side and API skills. Node.js and Python services, RESTful APIs in Express or FastAPI, microservice design, and the ability to review and challenge an architecture.
- Cloud and DevOps.
AWS (preferred), Azure or GCP, including managed AI/ML services, plus Docker, Kubernetes, Git workflows and CI/CD.
- Security and privacy mindset, with experience protecting sensitive data and owning risk trade-offs in a regulated or data-sensitive domain.
- Collaboration and mentoring. You work well with product managers, data scientists and UX designers, and you grow the people around you.
- Ownership and speed. A strong vision and the ability to deliver quickly and take complete ownership in a fast-moving startup environment.
Good to Have
- Experience in fintech, lending or payments, including credit decisioning, fraud models or integrating with banks, NBFCs and payment gateways.
- Knowledge of regulatory expectations for AI in financial services, including RBI guidelines and data localisation.
- Experience with real-time inference, feature stores and A/B testing of models.
- Open-source contributions, publications, conference talks or relevant certifications (for example AWS Machine Learning Specialty).
📌 AI/ML Architect (7-10 Years) (Gurugram)
🏢 Skeps
📍 Gurugram