New opportunity

Senior ML & AI Technical Solutions Engineer

Databricks · Bengaluru, India

About this role

As a Senior ML and AI Technical Solutions Engineer at Databricks in Bengaluru, India, you will play a critical role by helping customers debug and maintain stable GenAI and ML workloads with AI agent systems using the Databricks platform. You will develop end-to-end product expertise across products such as Agent Bricks, Vector Search, and Model Serving, collaborating cross-functional with engineering and account teams to ensure optimal customer success.

Responsibilities

Act as a senior technical solution expert for complex issues spanning data pipelines, ML pipelines, and AI applications using distributed systems.
Analyse and troubleshoot production workloads at the code level, optimising for performance, reliability, latency, and cost.
Diagnose and support Machine Learning and Large Language Model deployments including real-time and batch inference, autoscaling, monitoring, logging, and alerting.
Guide customers in leveraging Databricks AI for generative AI use cases, LLMs, MCP, AI agents, RAG, vector embeddings, semantic search, and prompt engineering.
Collaborate with internal teams to influence product roadmaps, improvements, and business growth.
Contribute technical documentation and wikis to share productionization expertise internally and externally.

Required qualifications

8+ years of experience designing, building, and scaling Data, Machine Learning, and AI systems on-premises and in the cloud using Python, Scala, and Java in production environments.
Expertise in Machine Learning and/or generative AI, along with data engineering necessary for orchestrating end-to-end ML training pipelines with Apache Spark.
Subject Matter Expert knowledge in feature engineering, ML frameworks, model training, monitoring, drift detection, and retraining strategies.
Experience working with algorithms, deep learning, and NLP techniques.
Prior experience building, designing, or troubleshooting LLM-based Generative AI applications and agentic frameworks such as LangChain and LangGraph.
Comprehensive knowledge of MLOps and LLMOps with expertise in model evaluation, scoring, ranking, optimisation, training, validation, and packaging.
Bachelor’s or Master’s degree in Computer Science, Engineering, or a related field (or equivalent experience).

Preferred qualifications

Experience developing agent skills, plugins, and debugging with native AI capabilities.
Prior support or customer-facing experience, along with a background in Data Scientist, ML Engineer, or AI Engineer roles.
Familiarity with cloud platforms (AWS, Azure, GCP) and Databricks.
Professional certifications.

Benefits

Comprehensive benefits and perks designed to meet the needs of employees in the region.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

PythonScalaJavaApache SparkLangChainLangGraphMLOpsLLMOpsVector SearchGenerative AIMachine Learning

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

Production Troubleshooting & Code-Level Optimization
Analyze and debug complex issues across distributed data pipelines, ML applications, and AI workloads at the code level, optimizing for performance, reliability, latency, and cost efficiency on the Databricks platform.
Generative AI & Agentic System Support
Guide customers through designing, deploying, and troubleshooting LLM-based applications, agentic frameworks (such as LangChain and LangGraph), vector search, retrieval systems, context orchestration, and prompt engineering.
MLOps & Lifecycle Observability
Serve as a subject matter expert on model monitoring, drift detection, experiment tracking, model registry, validation, and lifecycle observability for real-time and batch inference workloads.
Cross-Functional Product Influence
Collaborate closely with internal engineering and account teams by feeding customer feedback into product roadmaps, contributing to internal/external technical documentation, and sharing productionization insights.

Likely challenges to think through

Navigating High-Complexity GenAI and Agentic Failures
Troubleshooting non-deterministic LLM behavior, prompt drift, context window limits, and agentic loop errors under tight customer pressure requires deep domain intuition. Prepare by practicing systematic debugging of RAG architectures, multi-agent frameworks like LangGraph, and vector database performance bottlenecks.
Balancing Deep Code-Level Root Cause Analysis with Customer Communication
Rapidly translating intricate, low-level distributed system failures in Python, Scala, or Java into clear, empathetic guidance for enterprise clients. Prepare by refining structured technical communication skills to explain complex cloud infrastructure and MLOps trade-offs to diverse technical stakeholders.

Practice your answers

Technical practice

Troubleshooting Distributed Spark Pipelines

What to coverDiscuss how you approach diagnosing memory overhead issues, executor OOM errors, or shuffle bottlenecks in large-scale Apache Spark data pipelines. Emphasize your methodology for code-level profiling, data skew identification, and performance optimization.

LLM Inference Scaling and Latency Optimization

What to coverExplain the architectural considerations when supporting real-time versus batch inference for Large Language Models. Address strategies for managing autoscaling, concurrency limits, token throughput, and latency reduction in production deployments.

Context Orchestration and RAG System Design

What to coverDescribe how you optimize context windows, memory management, and retrieval systems in RAG or agentic RAG architectures. Discuss trade-offs between chunking strategies, embedding models, and semantic search accuracy.

Agentic Frameworks and Tool Integration

What to coverWalk through how frameworks like LangChain or LangGraph manage state, tool calling, and multi-step agent reasoning. Explain how you debug infinite loops or context degradation in agentic workflows.

Model Monitoring, Drift Detection, and Retraining

What to coverOutline your approach to detecting data and concept drift in machine learning models deployed in production. Discuss how you establish feature stores, metric tracking, model registries, and automated retraining strategies.

Vector Databases and Semantic Search Performance

What to coverExplain the technical trade-offs of using vector search databases (such as Vector Search or Lakebase databases) at scale. Address indexing methods (e.g., HNSW vs. IVF), distance metrics, and query latency considerations.

Code-Level Debugging in Python, Scala, or Java

What to coverDescribe a scenario where you had to debug a complex performance bottleneck or concurrency issue in a production system written in Python, Scala, or Java. Highlight the profiling tools and diagnostic steps you utilized.

LLMOps Evaluation and Lifecycle Observability

What to coverDiscuss the frameworks and metrics you use for comprehensive LLMOps lifecycle observability, including experiment tracking, output evaluation, labelling, tracing, and scoring LLM applications.

Behavioural practice

Resolving High-Pressure Production Outages

What to coverDescribe a situation where you had to troubleshoot a critical, high-impact failure in a distributed data or machine learning pipeline under tight time constraints. Use the Situation-Action-Result format to explain how you isolated the root cause and restored stability.

Guiding Non-Expert Customers Through Complex Technical Issues

What to coverShare an example of how you communicated an intricate technical solution or architectural limitation to a stakeholder or customer with varying technical depth. Detail your approach to empathetic listening, structured explanation, and ensuring successful resolution.

Cross-Functional Collaboration with Engineering Teams

What to coverProvide an instance where you worked directly with core engineering or product teams to report a recurring customer friction point or bug. Explain how you translated field observations into actionable product improvements or roadmap feedback.

Adapting to Rapidly Evolving Generative AI Technologies

What to coverDescribe how you stay ahead of fast-paced advancements in GenAI, MLOps, and agentic systems, and share an instance where you rapidly adopted a new tool or framework to improve operational efficiency or solve a customer challenge.

Sharpen your knowledge

Topics to prepare

Distributed Systems and Performance Tuning
Revision Exercise: Review distributed cluster architectures, memory allocation models, shuffle bottlenecks, and JVM tuning parameters in high-scale environments. Relevance: Directly supports the requirement to analyze and troubleshoot production workloads at the code level, optimizing for performance, reliability, latency, and cost across distributed data and ML pipelines.
LLMOps and Model Lifecycle Observability
Revision Exercise: Study experiment tracking, model registries, versioning, trace evaluation, labelling, and deployment monitoring frameworks. Relevance: Directly maps to the core responsibility of diagnosing and supporting Large Language Model deployments, ensuring rigorous lifecycle observability, scoring, ranking, and validation.
Context Orchestration and RAG Architectures
Revision Exercise: Practice designing retrieval systems, vector embeddings, semantic search, prompt design, and memory management strategies for GenAI applications. Relevance: Aligns with guiding customers on generative AI use cases, vector search databases, AI agents, and agentic RAG workflows.
Feature Engineering and Drift Detection Strategies
Revision Exercise: Review methodologies for feature stores, statistical drift detection, model retraining strategies, and monitoring ML frameworks in production. Relevance: Fulfills the explicit need for SME knowledge in feature engineering, model training, monitoring, and proactive drift detection across machine learning pipelines.

Skills to revise

Apache Spark and Large-Scale Data Engineering
Revision Exercise: Write and optimize Spark data processing jobs in Python or Scala to handle large datasets and orchestrate end-to-end ML training pipelines. Relevance: Directly addresses the requirement for proficient data engineering necessary for training pipelines and managing large-scale workloads.
LangChain and LangGraph Agentic Frameworks
Revision Exercise: Build and debug a multi-agent application using LangChain or LangGraph that integrates tool calling, memory management, and retrieval-augmented generation. Relevance: Directly matches the requirement for familiarity with agentic frameworks used to construct AI agent systems on the Databricks platform.
Python, Scala, and Java Production Programming
Revision Exercise: Inspect, refactor, and profile production-grade code snippets in Python, Scala, or Java to isolate memory leaks, concurrency issues, or latency bottlenecks. Relevance: Supports the requirement for 8+ years of experience designing, building, and scaling systems in production environments using these core languages.

Your preparation checklist

Prepare Distributed Systems and Code Troubleshooting Examples
Compile concrete examples from your career demonstrating how you diagnosed and resolved complex performance bottlenecks, latency issues, or memory failures at the code level in Python, Scala, or Java across distributed environments.
Build a Sample GenAI and Agentic RAG Application Walkthrough
Develop a compact demonstration project using an agentic framework like LangChain or LangGraph that integrates vector search embeddings, semantic retrieval, and context orchestration to showcase your practical GenAI engineering capabilities.
Document MLOps and LLMOps Lifecycle Observability Expertise
Prepare a structured summary of your experience managing model training, feature engineering, drift detection, experiment tracking, model registry, and validation strategies for production machine learning workloads.
Synthesize Large-Scale Apache Spark and Data Pipeline Experience
Outline your background in orchestrating end-to-end ML training pipelines and processing large datasets using Apache Spark to substantiate your core data engineering competencies.
Formulate Thoughtful Employer Questions on Support Engineering and Product Roadmaps
Draft targeted questions about how the global support engineering organization collaborates with core product and engineering teams to shape roadmaps for tools like Agent Bricks, Vector Search, and Model Serving.

Availability has not yet been verified. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in India. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…