AI Engineer (Chandigarh)

AI Engineer (Chandigarh)

20 Nov
|
Nayberry
|
Chandigarh

20 Nov

Nayberry

Chandigarh

The role

We sell products and services online through performance marketing, and our ad and marketing videos are produced by an AI pipeline we're building. Python agents check raw footage, plan the cuts, edit, caption, score the result, and deliver it.

You'll own and extend this pipeline: writing the Python modules, building the scoring and quality gates, shipping the review UIs, and running multiple agent sessions in parallel. Feedback arrives as short, informal chat messages, and your job is to turn it into a noticeably better output within one or two iterations.

You don't need to have built this exact system before. We want an engineer who already ships real software with LLMs, picks up a current codebase quickly, and turns messy feedback into a better result in one or two iterations. We'll get you up to speed on our pipeline and how we work. The engineering judgment has to be yours already.

Location: Chandigarh (hybrid, 3 days per week in office)

Hours: US Eastern time

Pay: ₹6 lac to ₹12 lac per year

What you'll do

- Build and iterate Python modules that check, score, cut, caption and deliver edited marketing videos.
- Design automated quality gates: LLM and rule-based scoring with thresholds and automatic redo loops.
- Run multi-agent orchestration, where a master session coordinates parallel agent sessions on different parts of a job.
- Ship small web review UIs (Cut, Plan and Review pages) that reviewers use to approve cuts and leave feedback.
- Turn reviewer feedback into updates to rule files (Markdown/JSON) versioned in GitHub.
- Write clear async progress updates on what changed, why, and what the scores did.
- Work on landing pages and marketing automation.

Must-haves

- 1+ years of professional experience building and shipping software (not internships alone).
- Strong Python: you can read, debug and restructure code you didn't write,



including code an AI wrote.
- At least one multi-step LLM workflow you built yourself, with tool calls, chained steps or agents. A single chatbot doesn't count. Personal work counts if it's real. A tutorial clone doesn't.
- You've shipped real software, in a job or on your own, and owned it after launch.
- A public GitHub repo or live project you can walk us through. This is required (see How to apply).
- Clear written English. Feedback comes in chat and you'll write progress updates daily.
- You think on your feet: when something breaks or the feedback is vague, you work out the next step instead of waiting to be told.
- You build things on your own: side projects, hackathons, or tools you made because you wanted them.
- Hybrid in Chandigarh (3 days a week in office) and US Eastern hours.

Nice to have

- Output evaluation or scoring: a judge model, test sets, guardrails or retry logic.
- A small web UI you've shipped, in any stack.
- LangGraph, CrewAI, the OpenAI or Anthropic tool-use APIs, or your own orchestration code.
- RAG with citations or guardrails.
- ffmpeg, MoviePy, Whisper or other transcription, caption formats such as SRT/VTT.
- Experience with AI coding tools such as Cursor, Claude Code or Codex, as long as you understand what they produce.
- A feel for what makes a short-form ad work (hooks, pacing, a clear CTA), or the keenness to learn it from our marketers.
- Landing pages or marketing automation, useful, not required.

What you'll pick up here





Our pipeline and scoring setup, our review tools and rule files, and what makes a good short-form ad for our products.

How to apply

Apply with your resume and a link to a public GitHub repo or live project where you built something with an LLM beyond a basic chatbot: a multi-step workflow, an agent, a RAG system, or a serious tool. Name the file or module you wrote that you're proudest of. A live product link, or a short walkthrough video work you can't share publicly, also works. Applications without a working link will not be reviewed. We reply to everyone who sends a working project link. If your link checks out, we'll ask you to explain your project in writing before any call. The job runs on written communication, so that's part of how we hire. If your experience doesn't match this list exactly but you've shipped real software and you learn fast, apply anyway.

The process

Short written exchange about your project, then a small build exercise in the pipeline pattern (you'll get a real clip, its scoring output, and a reviewer note, and improve one step), then a final conversation.

Pay: ₹600,000.00 - ₹1,200,000.00 per year

Benefits:

- Commuter assistance
- Internet reimbursement

Application Question(s):

- This role is hybrid in Chandigarh, 3 days a week in the office, and the working hours are US Eastern time. Can you commit to both?
- Link to a public GitHub repo or live project where you built something with an LLM beyond a basic chatbot: a multi-step workflow, an agent, a RAG system, or a serious tool. Name the file or module you wrote. Personal work counts. A live product link or a short walkthrough video of private work also works. Applications without a working link won't be reviewed.

Work Location: Hybrid remote in Chandigarh, Chandigarh (Chandigarh)

📌 AI Engineer (Chandigarh)
🏢 Nayberry
📍 Chandigarh

Reply to this offer

Impress this employer describing Your skills and abilities, fill out the form below and leave Your personal touch in the presentation letter.

Subscribe to this job alert:

Get the latest job offers by email for: ai engineer (chandigarh) / chandigarh