Artificial Intelligence Engineer

Aelum Consulting - ServiceNow Premier Partner; Dadri; On-site; INR 2,800,000 - 4,200,000; Full time

Job summary

Aelum Consulting - ServiceNow Premier Partner in Dadri is seeking an AI Solution Engineer — Agentic & Generative AI to be the technical face of our AI practice. You’ll translate client needs into production-ready architectures and then lead a small engineering team, staying hands-on with GenAI in Python.

You will build RAG pipelines, LLM orchestration, tool calls, and governance-aware patterns across enterprise platforms, while driving demos, PoCs and customer value.

Qualifications

• 5–8 years in software/AI engineering with GenAI/agentic delivery in production or PoC.
• Strong Python and GenAI stack: RAG, LLM orchestration, agentic patterns, prompt engineering, evaluation.
• Experience integrating LLMs with enterprise systems and cloud AI services (Azure OpenAI).
• Solid grasp of NLU/intent classification, vector search, embeddings and retrieval design.
• Excellent communication and client-facing presence.
• Proven ability to solution, estimate and lead.

Responsibilities

• Solutioning & architecture: translate client problems into target architectures with clear trade-offs.
• Hands-on build: productionise GenAI components, RBAC, and governance-friendly patterns.
• Team leadership: mentor a Python engineering team and drive delivery quality.
• Contribute to proposals with architecture narratives and ROI framing.

Skills

GenAI / agentic AI

Python

LLM orchestration

RAG pipelines

Prompt engineering

Enterprise integration

RBAC governance

Client-facing communication

Tools

Azure OpenAI

Python tooling

Job description

AI Solution Engineer — Agentic & Generative AI

About the role

You’ll be the technical face of our AI practice — the person who turns a client’s ambiguous "we want an AI assistant" into a credible, grounded, production-ready solution and then leads the team that builds it. Our work centers on enterprise conversational and agentic AI assistants that combine hybrid RAG (structured + document retrieval), LLM orchestration and query planning, deterministic business reasoning, and permission-aware, auditable design. Recent builds include agentic incident-management agents and multi-domain governance and contract-intelligence assistants that reason over enterprise records and documents through a hybrid architecture spanning approved enterprise platforms and Azure OpenAI.

This is a hybrid engineer–architect–lead role. You’ll solution and pitch in pre‑sales, stay hands‑on in Python, and grow a small engineering team.

What you’ll do

Solutioning & architecture

• Translate client problems and RFPs into target architectures, weighing platform‑native vs. hybrid orchestration vs. external‑assistant options and recommending with clear trade‑offs on cost, security, latency and governance.
• Design agentic and GenAI patterns end to end: intent/entity extraction, query planning, hybrid structured + document retrieval, deterministic calculation layers, grounded synthesis with citations, and human‑in‑the‑loop action execution.
• Apply "deterministic where you can, generative where you must" — keeping governed values (amounts, statuses, scores) computed or retrieved, never guessed.
• Lead discovery: ask the right business, process and data questions before positioning a solution, and convert answers into scope, assumptions and effort.
• Build and deliver compelling demos and PoCs; present architecture and value convincingly to both technical and business stakeholders.
• Contribute to proposals — effort estimation, architecture narrative, licensing assumptions, acceptance criteria and ROI framing.

Hands‑on build (GenAI / agentic)

• Build and productionise GenAI/agentic components in Python: RAG pipelines, LLM orchestration, tool/function calling, prompt and evaluation harnesses, retrieval and ranking, and API integration.
• Integrate LLMs (Azure OpenAI and other enterprise‑approved models) with enterprise systems via approved APIs, ensuring RBAC, data‑residency and audit constraints are respected.
• Stand up LLMOps: prompt/model versioning, golden test sets, evaluation, telemetry, cost monitoring and release gates.

Team leadership

• Lead and mentor a Python engineering team — set technical direction, review designs and code, unblock, and raise the delivery bar.
• Own quality: grounding accuracy, RBAC/no‑leakage, hallucination control, and reliable action execution.

What you’ll bring (must‑have)

• 5–8 years in software/AI engineering, with recent, demonstrable delivery of GenAI / agentic AI solutions in production or advanced PoC.
• Strong Python and the modern GenAI stack: RAG, LLM orchestration, agentic patterns (planning, tool use, multi‑step reasoning), prompt engineering, and evaluation.
• Practical experience integrating LLMs with enterprise systems and cloud AI services (Azure OpenAI or equivalent).
• Solid grasp of NLU/intent classification, vector search/embeddings, and structured‑vs‑unstructured retrieval design.
• Excellent communication and client‑facing presence — able to run discovery, present architecture, and demo credibly to mixed audiences.
• Proven ability to solution, estimate, and lead — you’ve owned technical scope and guided other engineers.
• Experience integrating AI into enterprise workflow platforms via APIs, events and middleware, with respect for platform RBAC, governance and audit controls.

Nice to have

• Experience with enterprise AI governance / responsible‑AI controls (RBAC‑aware retrieval, prompt‑injection mitigation, auditability, NIST AI RMF / OWASP LLM concepts).
• Exposure to open‑weight / self‑hosted LLMs and model portability strategies.
• Domain exposure to enterprise workflows (ITSM, CLM, vendor/risk, finance) that these assistants serve.
• Pre‑sales / consulting background in a partner or SI environment.