- Location
- Mumbai, India
- Workplace
- Onsite
- Type
- Full-time
- Department
- Engineering
- Source
- Keka
Description
AI Product & Automation Engineer
MetaGO | Turn bold ideas into working systems.
Reports to: Zaheer, Co-founder & CTO
Role: Individual contributor
Location / work arrangement: Mumbai · On-site
Employment type: Full-time
One builder. End-to-end ownership.
Some ideas arrive as a sentence. You can see the questions to answer, the first version to build, and the path to making it useful.
MetaGO is hiring one hands-on engineer to work directly with our Co-founder & CTO, Zaheer, and turn early ideas into working AI products, internal tools, and automations.
You will help shape the problem, work with the departments it affects, build the solution, and own it through launch and improvement. This is a role for someone who enjoys both writing the code and understanding the business behind it.
What you will own
- Turn ideas into clear, practical builds. Take an early idea, ask focused questions, identify assumptions, and define the smallest useful version.
- Build and ship independently. Develop applications, integrations, agents, and workflows without needing a separate development team to deliver each project.
- Create AI systems that do useful work. Give agents appropriate tools, context, and memory; evaluate their outputs and build sensible checks and human review where needed.
- Connect our tools and processes. Build integrations using APIs, webhooks, databases, workflow platforms, and scheduled jobs.
- Work across departments. Partner with teams such as sales, marketing, operations, HR, and finance to understand workflows, gather inputs, resolve dependencies, and agree on outcomes.
- Own adoption and reliability. Test with users, deploy, document, monitor, fix failures, and improve what you launch.
- Communicate with clarity. Explain trade-offs, flag blockers early, and give stakeholders a realistic view of progress and scope.
What you should already be able to show
- A working application, agent, or automation you personally built and shipped, with a clear explanation of your contribution.
- Strong coding ability in Python and/or JavaScript/TypeScript, plus practical experience with APIs, databases, authentication, and deployment.
- A multi-step workflow connecting several tools and running on a schedule or trigger.
- Experience building with LLMs, including tool calling, retrieval or persistent context, and checking output quality.
- The judgment to choose between ordinary software, a simple automation, and an AI agent based on the problem.
- Evidence that you can take an unclear request, work directly with stakeholders, and deliver something people use.
- The ability to diagnose failures and explain what you changed to make a system more reliable.
Relevant tools may include n8n, Make, Zapier, Pipedream, LangGraph, CrewAI, AutoGen, MCP, pgvector, Pinecone, Weaviate, Chroma, and Playwright. Equivalent tools and custom implementations are welcome. We care about what you can build; this is not a requirement to know every tool listed.
How you will work
You will be an individual contributor with broad ownership and direct access to Zaheer. You will work independently on delivery and closely with colleagues for context, access, feedback, and adoption.
You and Zaheer will agree on priorities and success criteria. You will decide how to move the work forward, communicate trade-offs, and keep the relevant people involved. We value focused questions, practical decisions, and steady progress.
What success looks like
- Useful first versions reach users quickly.
- Solutions improve a measurable business outcome, such as time saved, fewer errors, or faster turnaround.
- Stakeholders understand the solution and use it in their work.
- Systems stay dependable after the demo, with clear documentation and visibility into failures.
How to apply
Send a link to something you have built: a working product, repository, agent, workflow, or short demo. Include a few lines explaining the problem, what you personally owned, and what happened after launch.
Personal, open-source, freelance, and professional work are all welcome. If your work is private, a walkthrough using anonymized material is fine.
Shortlisted candidates will discuss a build with Zaheer and receive a small, scoped practical challenge with a 48-hour submission window. We will assess your decisions, the working result, and what you would improve next.
Apply: [email protected]
Show us something you’ve built.
Skills
PythonJavaScriptTypeScript