Company Description Linxus InfoTech is a cloud infrastructure software company focused on giving engineering and DevOps teams clarity, control, and confidence over their cloud environments. The company builds tools that simplify complex cloud setups, where resources are spread across services, regions, and accounts, and visibility and security are often handled manually. Its flagship product, InfraSync, enables teams to discover, sync, and secure their AWS infrastructure from a single, unified interface.

This product is the foundation of a broader vision to make cloud infrastructure effortless to understand and manage. Founded in India, Linxus InfoTech designs solutions for distributed teams around the world.

Role Description We are looking for an AI Economics & Agent Engineer to build and operate enterprise-grade AI solutions across AWS, Microsoft Azure and Google Cloud.

The ideal candidate will have strong hands-on experience with cloud platforms and their LLM/GenAI services, particularly Amazon Bedrock, Azure OpenAI / Azure AI, and Google Vertex AI, combined with experience building AI agents, LLM orchestration, RAG systems and cloud cost optimization.

You will work on a platform that helps enterprises understand, optimize and govern their AI consumption including token usage, model selection, AI infrastructure cost, agent efficiency, forecasting and business value.

Qualifications

Experience: 0 - 2 years

Employment: Full-time

Location: Remote / Hybrid

Domain: Cloud, GenAI, AI Agents, FinOps

Level: Mid Engineer / Senior Engineer

About the Role

Key Responsibilities

- Multi-Cloud LLM Engineering

Build and integrate enterprise GenAI solutions across:
- AWS Amazon Bedrock
- Microsoft Azure OpenAI / Azure AI
- Google Cloud Vertex AI / Gemini
- Open-source and third-party LLMs where appropriate

Evaluate models based on

- Input/output token cost
- Latency
- Quality
- Context-window requirements
- Reasoning capability
- Reliability
- Data residency
- Security/compliance
- Business requirements

Design model-routing strategies so workloads use the most appropriate model rather than automatically using the most expensive model.
- AI Agent Engineering

Design and develop production-grade AI agents using frameworks such as:
- LangGraph
- LangChain
- Microsoft Agent Framework / Agents SDK
- Python
- REST APIs
- MCP / tool integrations

Build agents capable of

- Investigating cloud costs
- Detecting anomalies
- Explaining cost increases
- Recommending optimization
- Performing policy checks
- Generating reports
- Interacting with enterprise APIs
- Taking actions with approval workflows

Experience with multi-agent systems and human-in-the-loop workflows is highly desirable.
- AI Economics & FinOps

Build capabilities to measure and optimize AI economics.

Responsibilities include

- Token consumption tracking
- Cost per model/request/user/team/application
- AI workload allocation
- Cost forecasting
- Budget monitoring
- Cost anomaly detection
- Model right-sizing
- Prompt optimization
- Context optimization




- Caching strategies
- Batch processing
- Cost-aware agent design
- Chargeback/showback The objective is not simply to reduce token cost, but to optimize cost per successful business outcome.
- RAG & Enterprise Knowledge

Build enterprise RAG solutions using:
- Vector databases
- Embeddings
- Document ingestion pipelines
- Semantic search
- Hybrid search
- Metadata filtering
- Retrieval optimization
- Context compression

Integrate enterprise sources such as:
- SharePoint
- Confluence
- ServiceNow
- Jira
- ERP systems
- Cloud billing systems
- Internal APIs
- Documentation repositories
- Cloud & Infrastructure

Strong cloud engineering experience is required.

The candidate should be comfortable with:

AWS

- Bedrock
- S3
- Lambda
- ECS/EKS
- CloudWatch
- IAM
- Cost Explorer
- CUR / billing data
- API Gateway
- EventBridge
- DynamoDB/RDS

Azure

- Azure OpenAI
- Azure AI Foundry
- Azure Monitor
- Azure Cost Management
- Entra ID
- Azure Functions
- AKS
- Storage

GCP

- Vertex AI
- Gemini
- Cloud Run
- GKE
- Cloud Storage
- BigQuery
- Cloud Monitoring
- Billing/FinOps APIs

You do not need to be an expert in every service, but strong hands-on experience in at least one cloud and practical exposure to the other two is expected.
- Microsoft 365 / Copilot Integration

Experience with the Microsoft AI ecosystem is highly valuable.

Knowledge of

- Microsoft 365 Copilot
- Copilot Studio
- Custom Engine Agents
- Microsoft Agents SDK
- Microsoft Graph
- Entra ID
- Teams
- SharePoint
- Enterprise connectors is preferred.

The role will involve designing custom enterprise agents that can work alongside the Microsoft 365 ecosystem.
- AI Governance & Security

Implement enterprise guardrails around AI consumption and usage.

Examples

- Token/cost thresholds
- Budget limits
- Model access policies
- PII/data protection
- Prompt injection protection
- Authentication/authorization
- Audit logging
- Human approval
- Model usage policies
- Responsible AI controls
- Data residency requirements

Experience with AWS Bedrock Guardrails, Azure AI safety capabilities or equivalent is a plus.

Required Technical Skills

Mandatory

- 3+ years of cloud engineering experience
- Solid Python development
- Hands-on experience with at least one major cloud platform
- Hands-on experience with LLM/GenAI services
- Experience with AWS Bedrock OR Azure OpenAI/Azure AI OR Google Vertex AI
- Understanding of:
- Tokens
- Context windows
- Embeddings
- Prompt engineering
- RAG
- LLM APIs
- Model selection
- AI inference economics
- Experience building REST/API-based integrations
- Docker and Kubernetes fundamentals
- Git and CI/CD




- Cloud IAM/security fundamentals

Strongly Preferred

- Experience with AWS + Azure + GCP
- Amazon Bedrock
- Azure OpenAI / Azure AI Foundry
- Google Vertex AI / Gemini
- LangGraph
- LangChain
- MCP
- AI agent development
- Multi-agent architecture
- Vector databases
- Microsoft 365 Copilot / Copilot Studio
- FinOps / Cloud Financial Management
- Cloud billing APIs
- Power BI
- Terraform / Infrastructure as Code
- AI observability
- LLM evaluation and benchmarking

Ideal Candidate

We are specifically looking for someone who can answer questions such as:

«"Our AI spend increased by 25% this month. Which applications, teams, models and workflows caused it?"»

and then build an agent that can:

Detect → Investigate → Explain → Recommend → Get Approval → Execute The candidate should be comfortable moving between:

Cloud infrastructure → LLMs → AI agents → Data → FinOps → Business value rather than working exclusively in one area.

Example Technologies

Cloud: AWS | Azure | GCP

LLM: Amazon Bedrock | Azure OpenAI | Azure AI | Vertex AI/Gemini | Open-source LLMs

Agent Frameworks: LangGraph | LangChain | Microsoft Agents SDK | MCP

Languages: Python | TypeScript/JavaScript

Infrastructure: Docker | Kubernetes | Terraform

Data: PostgreSQL | DynamoDB | BigQuery | Vector DBs

Analytics: Power BI | Cloud billing APIs | Custom dashboards

DevOps: GitHub/GitLab | CI/CD | Monitoring | CloudWatch/Azure Monitor/GCP Monitoring

What Success Looks Like

Within the first few months, you should be able to:

- Build a working multi-cloud LLM integration
- Track AI/model/token consumption
- Identify expensive AI workloads
- Implement model routing and right-sizing
- Reduce unnecessary context/token consumption
- Build an AI economics agent
- Connect the agent to enterprise data and APIs
- Implement governance and approval controls
- Produce actionable AI cost/ROI dashboards
- Deploy the solution securely in a production cloud environment

Nice to Have

Candidates with experience in any of the following will stand out:
- FinOps certification
- AWS/Azure/GCP certification
- Experience managing enterprise AI platforms
- LLM cost optimization
- AI observability platforms
- OpenTelemetry
- Prompt/token telemetry
- LLM evaluation frameworks
- Enterprise Copilot implementations
- Cloud cost optimization
- SaaS platform development
- Production AI agent deployments

Interview Focus The interview will be heavily hands-on and architecture-oriented.

Candidates should expect discussions or practical exercises around:

- Designing a multi-cloud LLM architecture
- AWS Bedrock model selection and cost optimization
- Azure OpenAI vs Bedrock vs Vertex AI
- Designing an AI cost monitoring system
- Building a LangGraph agent
- RAG architecture
- Token/cost optimization
- Model routing
- AI governance
- Cloud FinOps
- Production deployment and security

We are looking for builders, not candidates who have only experimented with ChatGPT or basic prompt engineering.

📌 AgenticAI Engineer (Pune)
🏢 Linxus InfoTech Private
📍 Pune