Forward Deployed Engineer

What you'll do

Sit with whichever business unit or founder you're deployed to, understand the real problem, and define a clear technical plan often before anyone else has scoped it.

Build and ship production code, prototypes, and everything between, using whatever the problem demands: data pipelines, LLM orchestration across providers, backend services, internal tools, deployment on AWS.

Decide, every time, whether AI/LLMs are actually the right tool for the problem and say so plainly when they aren't.

Present your work directly to the stakeholders (founders, business unit heads, functional teams) who asked for it. You own the narrative, not just the code

Train the team you're embedded with so what you built keeps running without you

Contribute reusable platforms, frameworks, and standards back to the AI Initiatives COE so the next deployment moves faster than the last

What makes you stand out

You've used AI coding tools (Claude Code, Cursor, or similar) as your daily driver for a meaningful stretch and not something you tried once

You orchestrate agents, not just prompt them. You build with skills, MCPs, and multi-step automation, and it's muscle memory

You have a public or demonstrable artifact, a shipped project, a hackathon win, a repo, something you can point to and say "I built this, end to end"

You've shipped something into a real workflow, even an informal one, that's still running without you

You're comfortable with zero spec and being told "just figure it out"

You're a fast generalist. You'd rather learn lease accounting, hospitality operations, or campus administration on the fly than wait for a briefing document

Founder mindset: you've taken something from a blank page to working, on your own hands, at least once

Comfortable being the only technical person in a room of non-technical stakeholders, and can explain your work in their language

Who should apply

This isn't limited to computer science grads or people already carrying an "AI Engineer" title. We're open to anyone, from any educational or professional background, at any stage: interns, fresh graduates, or experienced professionals from any discipline who've been wanting to make the move into AI and can show they've already started. What matters is what you've built, not what your degree or job title says.

How to apply:

The Build TrackWe're skipping the resume screen and the standard interview funnel for this role. Instead, you pick a problem statement below (or propose your own under the Open Track), build a working solution against it, and submit it. Strong submissions get called in directly for a live panel. No aptitude test, no group discussion in between.

What to submit, to

ai@divyasree.com

• A public GitHub repository with your code. Include a clear README covering what it does, how to run it, and how it's set up.
• A 5-minute pitch video (unlisted link is fine): Lead with the problem, then your solution, then a live demo of it actually working. Spend the most time on the AI-native part: if your project touches agents, orchestration, or LLM reasoning, that's what we want to see in action. This is optional but recommended.
• A short written note (half a page is enough) covering: what tradeoffs you made and why, where this would break at scale or on real data, and where you did or didn't use AI and why.

Send the repo link, video link, and note in a single email to

ai@divyasree.com

, with the track name in the subject line. Feel free to also share links to anything else you've built that you'd like us to see (other repos, side projects, live products, anything that shows how you think and build)

What we're evaluating:

AI judgment:

Did AI/LLMs genuinely add value here, or were they forced in because the brief mentions AI?

Speed-to-signal:

How much real progress did you make in the time given, versus over-engineering a narrow slice?

Failure awareness:

Do you know where this breaks, and have you said so honestly?

Communication:

Can you explain this to a non-technical founder or business head in plain language?

Scope instinct:

Did you pick something narrow enough to actually finish?

What happens next:

Shortlisted submissions move to the one-hour, hands-on technical call described below. From there, strong candidates go straight to an offer conversation — there's no additional round.

Problem statement

Pick one, or propose your own under the Open Track. Every track uses the same submission format above (GitHub repo + README, 5-minute pitch video, short note).

Open Track

Propose and build your own idea for a real problem anywhere in the Group: real estate, hospitality, co-living, facilities, or asset management. A working prototype, plus the same note format as above.

What the technical call looks like

Shortlisted from a Build Track submission, or applying directly: the first technical call is one hour, hands-on, and runs entirely off your screen. We'll walk through what you built, then hand you a new, live, underspecified problem and watch how you scope it and start, not whether you finish it.

What you'll get

Broad exposure: you'll work across nearly every part of a diversified Group spanning real estate, hospitality, co-living, facilities, and education within your first year

Real ownership from week one, wherever you're deployed: your work is load-bearing, not a side project

Freedom to use whatever models, agents, frameworks, and tools work best for the problem in front of you

Direct access to founders and business unit leaders: your work is judged on impact, not process

A growing track record of shipped, adopted systems across the group, not just POCs

High-variance work by design: some deployments are two-day sprints, some become the system a whole business unit runs on

DivyaSree Group is an