New opportunity

Senior AI/ML Engineer

Chapter247 Infotech · Indore, India

About this role

Chapter247 Infotech seeks a Senior AI/ML Engineer in Indore to own AI solutions across the software development lifecycle. The role involves understanding business problems, working with clients and stakeholders, translating requirements into scalable designs, and guiding solutions through development, deployment and continuous improvement.

Responsibilities

Design scalable backend systems, AI workflows and architectures; develop and optimize applications using LLMs, RAG pipelines, OCR and agentic AI frameworks.
Evaluate and benchmark models, embeddings, prompts, retrieval strategies and frameworks. Run experiments and proofs of concept for new AI capabilities.
Build secure, reliable production systems with guardrails, validation, observability and performance optimization. Troubleshoot complex issues and implement lasting solutions.
Integrate third-party APIs, cloud platforms and external AI providers. Contribute to architecture decisions, estimates, coding standards, testing and engineering practices; communicate project updates and mentor junior developers.

Qualifications and skills:

The posting specifies 4–6 years of experience and lists a bachelor's degree in its structured requirements. Strong Python and backend experience is sought, including FastAPI, Django REST or Flask RESTful; Pydantic, SQLAlchemy, asynchronous programming, authentication and scalable REST APIs.
Practical experience with generative and agentic AI, LLMs, prompt engineering, RAG, OCR, tool calling, memory and context management, and autonomous workflows. Experience with NLP, embeddings, semantic search, model training, TensorFlow or PyTorch, Hugging Face and vector databases is also sought.
Familiarity with AI safety measures such as prompt-injection prevention, hallucination mitigation, content filtering and secure tool execution; distributed workflows and messaging; databases; cloud services; Git, CI/CD, Docker, Terraform, Pytest, secure coding and API security. The role also calls for using AI-assisted development tools.

Preferred

Frontend experience with JavaScript, HTML5, CSS and React.

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

Skills for this role

PythonFastAPIDjango RESTFlask RESTfulPydanticSQLAlchemyAsynchronous programmingREST API designGenerative AIAgentic AILarge language modelsPrompt engineeringRetrieval-augmented generationOCRAI agentsTool callingNLPEmbeddingsSemantic searchModel trainingTensorFlowPyTorchHugging FacePineconeWeaviateQdrantChromaDBFAISSTemporalLangGraph

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

Own the solution lifecycle
The posting calls for understanding business problems with clients and stakeholders, translating requirements into scalable designs, and driving solutions through deployment and improvement. Prepare a concise account of how you make and communicate design decisions across those stages.
Build AI applications on scalable backends
The role develops and optimizes LLM, RAG, OCR and agentic applications, alongside Python backend systems and APIs. Prepare to explain how an AI workflow fits into a maintainable backend architecture.
Evaluate before committing to an approach
Responsibilities include benchmarking models, embeddings, prompts and retrieval strategies, plus running experiments and proofs of concept. Prepare to describe evaluation criteria, comparisons and how results would inform an implementation choice.
Make AI systems production-ready
The posting emphasizes guardrails, validation, observability, security, performance optimization and lasting fixes to complex issues. Prepare to explain how you would detect, diagnose and address failures after deployment.
Provide technical leadership
The engineer contributes to architecture decisions, estimates, coding standards and testing; integrates external services; communicates updates; and mentors junior developers. Prepare examples of explaining trade-offs and giving actionable technical guidance.

Likely challenges to think through

Balancing AI quality with system performance
Likely challenge, not a confirmed employer problem: a model or retrieval change may improve answer quality while making the application slower or harder to scale. Prepare a benchmark plan that compares output quality and performance against the business requirement.
Constraining agentic and external actions
Likely challenge, not a confirmed employer problem: tool calling and third-party integrations create useful workflows but also expose systems to prompt injection, unsafe execution and unreliable responses. Prepare to discuss permission boundaries, response validation, guardrails and observable tool behavior.
Taking experiments into production
Likely challenge, not a confirmed employer problem: a successful proof of concept may not yet be reliable or maintainable in a deployed backend. Prepare to distinguish experimental evidence from production readiness, including testing, deployment, monitoring and root-cause investigation.
Resolving ambiguous requirements
Likely challenge, not a confirmed employer problem: stakeholder goals may not specify the technical constraints needed for an architecture or estimate. Prepare a method for clarifying priorities, presenting options and communicating decisions to technical and non-technical audiences.

Practice your answers

Technical practice

How would you turn a business requirement into a scalable AI-backed REST API?

What to coverThe posting calls for requirements analysis, Python backend design, authentication and scalable REST APIs. Suggested answer focus: explain API boundaries, validation, asynchronous work where appropriate, and how you would test the design against the stated requirement.

How would you design and improve a RAG pipeline?

What to coverThe role includes RAG, embeddings, semantic search and vector databases. Suggested answer focus: discuss ingestion, retrieval and answer generation, then compare retrieval strategies using relevance, latency and grounded-answer quality.

How would you control an agent that calls tools across multiple steps?

What to coverThe posting names agentic workflows, tool calling and memory or context management. Suggested answer focus: describe tool permissions, context limits, step validation and when a simpler workflow would be more reliable.

How would you build and validate an OCR-based AI workflow?

What to coverDeveloping and optimizing OCR systems is a stated responsibility. Suggested answer focus: explain how you would check extraction quality, handle uncertain or malformed text, and prevent downstream AI responses from treating bad extraction as reliable evidence.

How would you benchmark models, prompts and retrieval approaches for a use case?

What to coverThe role explicitly includes experiments, proofs of concept and benchmarking of models, embeddings, prompts and retrieval strategies. Suggested answer focus: define representative test cases and compare quality, latency and cost before choosing an approach.

How would you defend an AI application against prompt injection and unsafe tool use?

What to coverThe posting lists prompt-injection prevention, guardrails, response validation, content filtering and secure tool execution. Suggested answer focus: distinguish untrusted input from authorized instructions, constrain tool access, and explain how you would test failure cases.

How would you diagnose a production AI system that becomes slow or gives unreliable answers?

What to coverObservability, performance optimization, root-cause analysis and lasting fixes are stated responsibilities. Suggested answer focus: isolate retrieval, model, API and infrastructure stages with traces or metrics, then explain how you would verify a fix without masking the underlying issue.

How would you integrate external AI services into a distributed backend workflow?

What to coverThe posting covers third-party APIs, cloud services, messaging, workflow orchestration and databases. Suggested answer focus: discuss timeouts, retries, failure recovery and data consistency, including the trade-off between synchronous requests and queued work.

When would you adapt an existing model rather than change the application pipeline?

What to coverThe skills list includes model training, TensorFlow, PyTorch and Hugging Face alongside prompt and retrieval engineering. Suggested answer focus: compare the evidence needed to justify training or adaptation against improving prompts, context or retrieval first.

How would you test and deploy a production AI service safely?

What to coverThe posting names Pytest, CI/CD, Docker, Terraform, secure coding and containerized deployment. Suggested answer focus: cover tests for backend behavior and AI failure cases, deployment checks, and a way to detect and address regressions after release.

Behavioural practice

Tell me about a time you clarified an ambiguous client or stakeholder request.

What to coverThe role involves understanding business problems and translating requirements into technical solutions. Suggested answer focus: use a real situation, the questions and decisions you drove, and the resulting agreement or change in direction.

Describe a time you took ownership of a solution beyond its initial implementation.

What to coverThe posting emphasizes ownership from design through deployment and continuous improvement. Suggested answer focus: give the situation, your actions across those stages, and an observable result without claiming work you did not do.

Tell me about an architecture decision where your team had competing priorities.

What to coverContributing to architecture decisions, estimates and technical discussions is part of the role. Suggested answer focus: explain the situation, how you weighed trade-offs and incorporated others' input, and what the decision achieved.

Describe a time you helped a junior developer work through a technical challenge.

What to coverThe posting explicitly calls for technical guidance, design recommendations and implementation support. Suggested answer focus: describe the situation, how you coached rather than simply took over, and the outcome for the work and developer.

Tell me about a time you communicated a technical risk or project update to a non-technical stakeholder.

What to coverThe role requires communicating solutions and updates to technical and non-technical stakeholders. Suggested answer focus: explain the situation, how you made the trade-off or risk understandable, and what action or decision followed.

Sharpen your knowledge

Topics to prepare

Scalable AI backend architecture
Suggested revision exercise: sketch an API-backed AI workflow and explain where authentication, asynchronous processing, database access and third-party integrations belong. The posting asks for scalable backend designs and architecture decisions.
RAG and model-selection trade-offs
Suggested revision exercise: compare two embedding or retrieval strategies on relevance, latency and failure cases, then explain how you would choose a model or prompt. The role explicitly involves benchmarking models, embeddings, prompts and retrieval strategies.
Agent workflows and orchestration
Suggested revision exercise: diagram a multi-step agent with tool calls, context management, retries and a stopping condition. The posting names agentic AI, autonomous workflows and distributed orchestration.
AI safety and API security
Suggested revision exercise: threat-model a retrieved document containing instructions to an agent; explain prompt-injection prevention, response validation and secure tool execution. These safeguards and OWASP/API security are explicitly listed.
Production reliability and observability
Suggested revision exercise: trace a hypothetical rise in latency and incorrect answers from symptoms to root cause, then propose measurements and a lasting fix. The role calls for observability, performance optimization and troubleshooting.
Requirements-to-design decisions
Suggested revision exercise: turn a short business request into clarifying questions, a design with trade-offs and an implementation estimate. The role spans stakeholder discussions, architectural decisions, project updates and technical guidance.

Skills to revise

Build a Python AI API
Suggested hands-on exercise: implement a small authenticated REST endpoint with FastAPI, Pydantic validation, SQLAlchemy persistence and an asynchronous external call. These are named backend skills; the exercise is preparation, not an advertised project.
Benchmark a RAG and OCR pipeline
Suggested hands-on exercise: extract text from sample documents, index embeddings in FAISS, and compare retrieval choices using a small set of questions and recorded results. OCR, RAG, vector databases and benchmarking appear in the posting.
Implement a guarded agent workflow
Suggested hands-on exercise: create a multi-step tool-calling flow with LangGraph or Temporal, constrain tool inputs and test an attempted prompt injection. The posting names agent frameworks, orchestration and secure tool execution.
Test and deploy a reliable service
Suggested hands-on exercise: add Pytest cases for invalid inputs and provider failures, containerize the service with Docker and outline a CI/CD check. Testing, Docker, CI/CD and production reliability are explicit.
Present an architecture review
Suggested practical exercise: give a brief design walkthrough for an AI service, state estimates and trade-offs, then explain a coding or testing recommendation as if mentoring a junior developer. The posting calls for stakeholder communication, estimation and mentoring.

Your preparation checklist

Map the role to your evidence
The posting calls for ownership across the software development lifecycle. Prepare a brief evidence sheet linking relevant work, where available, to requirements discovery, architecture, deployment, improvement and stakeholder communication; mark gaps as topics to discuss, not claims of experience.
Rehearse an end-to-end project walkthrough
Prepare one relevant AI application example, if available: the business problem, decisions you made, your contribution, measurable results and what you changed after release. Be ready to explain it to both technical and non-technical stakeholders.
Practice a scalable AI-system design
Sketch a hypothetical Python backend for an LLM application using RAG, OCR or tool calling. Explain API boundaries, data and context flow, authentication, asynchronous work, external-provider integration and scaling trade-offs. This is a preparation exercise, not a stated company project.
Build an evaluation and experiment plan
The role includes benchmarking models, embeddings, prompts and retrieval strategies. Prepare a compact comparison plan with test cases, quality and performance measures, a baseline, and a rule for deciding whether a proof of concept should proceed.
Prepare a production-readiness review
Create a checklist for prompt-injection prevention, secure tool execution, response validation, hallucination mitigation, observability and failure handling. Include one example of how you would investigate a production issue and verify a lasting fix.
Collect backend and delivery examples
Have concise examples, if available, of Python REST API design, testing, deployment and third-party integration. Be ready to discuss choices involving frameworks, databases, CI/CD, containers and secure coding without claiming experience you cannot substantiate.
Prepare leadership examples and employer questions
Outline one example each of an architecture decision, estimate or technical trade-off, and guidance to a junior developer, where applicable. Ask how this role works with clients, evaluates AI systems, handles production ownership and sets engineering standards.

Availability last checked: 4 Nov 2026. 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…