AI Software Engineer (India) | RQ-0029

Location :Navi Mumbai (Seawoods_
Work Model :Hybrid
Position Type:Full-time, Permanent
About Us
Founded in 1995, Zema Global Data Corporation empowers organizations to simplify complexity, reduce risk, and make faster, more confident decisions that drive measurable results. Over the past two years, Zema Global has accelerated its growth through strategic investment and acquisition to strengthen our global leadership. Together we’re helping our customers gain a Decisioning Advantage - one bold idea at a time. With a presence across global energy, commodity, and financial markets, Zema Global empowers customers to simplify complexity, reduce risk, and make faster, more confident decisions that drive measurable results.
At Zema Global, we Think Big, Make It Happen, and Win as One. We thrive on collaboration, creativity, and respect, united by a shared drive to innovate and deliver meaningful impact for our customers and communities. If you’re inspired by solving complex challenges and contributing to a culture that values purpose and performance, we invite you to join us.
Position Overview
We are hiring an AI Software Engineer with 4+ years of experience to build and run AI agents across our data and analytics platform. These agents will take on a growing range of work for customers and internal teams, using LLMs, tools and retrieval in production. You will build agents end to end and help operate them once they are live, monitoring their behavior, cost and quality (often called Agent Ops). You will work closely with senior engineers, data engineers and product managers, and ship to real customers.
What You'll Be Doing
Build agents
Asa the AI Software Engineer, you will design and build LLM agents that plan multi-step tasks and call tools, such as querying internal data APIs, running analytics or generating reports.
Rapidly prototype new agent ideas, put them in front of users early and iterate quickly based on feedback.
Define clear tool interfaces (names, schemas, descriptions) and connect agents to internal services, including through protocols like MCP.
Build retrieval (RAG) so agents ground answers in governed company data: chunking, embeddings, vector search and citations.
Write and iterate on system prompts, tool descriptions and guardrails.
Contribute to multi-agent and human-in-the-loop workflows where a task needs review or approval.
Run and monitor agents (AgentOps).
Instrument agents with tracing so every step, tool call and output can be inspected and replayed.
Track quality, latency, token cost, tool error rates and task success in dashboards and alerts.
Build and maintain evaluation sets; run evals before each release to catch regressions.
Investigate failures such as wrong tool calls, loops, hallucinations and timeouts, and ship fixes.
Keep agent actions traceable and auditable, in line with customers' governance and compliance needs.
Engineering
Write production Python services and APIs, and deploy them on cloud infrastructure with CI/CD.
Work with product and domain experts to turn customer problems into agent use cases.
Write clear documentation and join code reviews.
Essential
What You'll Bring
4+ years of professional experience in AI/ML or software engineering (internships and strong projects count).
Bachelor's degree in Computer Science, Engineering, Mathematics, Statistics or a related field, or equivalent experience.
Strong Python skills, including async code, typed data models (e.g. Pydantic) and REST APIs.
Hands-on experience building at least one LLM agent or tool-using application, at work or in a project you can show us.
Ability to prototype rapidly and iterate on ideas quickly: turning a rough idea into a working demo in days, then improving it.
Experience with LLM APIs (for example Anthropic, OpenAI or Azure OpenAI), function/tool calling and structured outputs.
Hands-on experience with LangGraph or a similar agent framework (such as CrewAI, AutoGen, LangChain or an agent SDK).
Understanding of how to evaluate LLM output: test sets, metrics and LLM-as-judge.
Working knowledge of SQL and Git.
Clear communication and a willingness to learn a new domain.
Desirable
Cloud agent platforms: AWS (Amazon Bedrock Agents or Bedrock AgentCore) or Azure (Azure AI Foundry Agent Service, Semantic Kernel or Microsoft Agent Framework).
Experience with agent observability or LLMOps tools such as LangSmith, Langfuse, Arize Phoenix, Weights & Biases or OpenTelemetry.
Building MCP servers or other tool integrations for agents.
Multi-agent systems, planning, memory or human-in-the-loop workflows.
Knowledge of agent security risks, such as prompt injection and over-broad tool permissions.
Deploying agents on AWS or Azure with Docker and Kubernetes.
Interest in or exposure to energy, commodities, trading or financial markets.
Time-series data and forecasting.
Vector databases (pgvector, Pinecone, Weaviate) and embedding models.
Why Join Zema Global?
ü Global Impact Shape the future of energy, commodities, and financial markets through innovative technology.
ü Innovation at Scale Work with cutting-edge software, advanced analytics, and complex data that solve real-world challenges.
ü Career Growth
Join a fast-growing global organization where your ideas, development, and career progression are valued.
ü Exceptional People Collaborate with talented colleagues and industry experts across North America, Europe, and Asia.
ü Meaningful Ownership Take ownership of your work, influence strategic initiatives, and make a measurable impact on our customers and business.
ü One Global Team Be part of a collaborative culture built on trust, respect, and our values: Think Big. Make It Happen. Win as One.
Please note: Visa sponsorship is not available for this position.
Equality, Diversity & Inclusion
Zema Global is committed to building an inclusive workplace where everyone can thrive. We encourage applications from all qualified candidates regardless of background or protected characteristic

Work arrangement
Hybrid