06 Aug
|
GIVA
|
Bengaluru
AI Engineer 1
Function Technology / Product & Engineering
Location Bengaluru, India (on-site)
Experience 0–2 years
Reports To Head of Data & AI Initiatives About Giva
GIVA is a large-scale jewellery brand with over 350 physical locations in India and abroad.
We build the tech that ties everything together: from customer-facing stores to the factories and logistics. Our software runs the entire business end to end, handling the warehouse and
ERP systems needed to scale fast. About the role
This is a software engineering job, you will build applications that use language models.
Most of your time goes into writing code, designing systems, and handling the cases where things break.
The main work is building agentic applications: programs that take a goal, break it into steps, call tools and data sources, and produce a result. You will work closely with the product and business side, who decide what gets built and why. What you will do
- Build product features that use language models. This covers the API calls, the prompts, MCPs, and the code around them.
- Set up retrieval over our own data so the model answers from real content. This means chunking, embeddings, vector search, and working out why the wrong document keeps coming back.
- Build agents that automate multi-step work. These call tools and recover when a step fails.
- Run the quality loop. Build test sets, run A/B tests, check whether a change actually improved anything, and tune prompts based on the results.
- Get your work into production and keep it running. Logging, monitoring, and watching what it costs.
- Compare foundation models from Anthropic/ Google/ OpenAI/ other open source models for your application, and pick the right one for each job. Accuracy, speed and cost matter. What we expect in 6-9 months
You build agentic applications end to end on your own. We decide what to build and why.
You handle the technical side: design, build, test, ship, fix, and keep it running.
This is a high bar for someone early in their career. We will support you, but the role does not involve being handed well-defined tickets for a year. What we are looking for
Projects you have built with AI - A side project, a hackathon build, or a tool you made for yourself. Something where you connected a model to real work. This is the first thing we look at.
Engineering fundamentals - You are comfortable in Python. You use Git, debug your own code, and read documentation. Other people should be able to work with the code you write.
Speed - You get a working version out early and improve it after feedback. Communication - You can explain what you built and why, including to people who are not engineers. You are comfortable saying you do not know something.
0-2 years of experience - Recent graduates with strong engineering/ML background. Valuable to have but not mandatory
- Experience with GenAI libraries such as LangChain, LangGraph, or Hugging Face
- Experience with vector databases such as Pinecone, Weaviate or Chroma
- Any experience running software in production: deployment, monitoring, or cost control
- A public GitHub, or a writeup of something you worked out
📌 AI Engineer 1 (Bengaluru)
🏢 GIVA
📍 Bengaluru