INDEPENDENT MINDS. SHARED AMBITION.
Question the
way things work.
Start with first principles. Reimagine with AI.
Build towards an outcome that matters.
Your curiosity belongs in the room where the next answer is being built. Begin your career at devx.
Challenge the assumption.
Make your thinking visible.
Own the customer outcome.
01 / THE COMPANY YOU HELP BUILD
BUILT FROM INDIA, FOR THE WORLD.Building for the AI world.
Solving for outcomes.
We’re an AI-native consulting firm reimagining customer interactions, business operations and enterprise architecture. We start by asking what the customer needs to change, then work backwards to the right solution.
Young talent brings fresh questions. Experienced practitioners bring context and judgment. At devx, those perspectives come together to turn ambitious ideas into useful work.
THE CULTURE, IN PRACTICE
Five principles.
A daily standard.
How you ask a question, share a decision or respond to a setback is how you shape the culture.
Understand what the customer is trying to change. Ask the extra question, surface the real obstacle and connect your work to their goals.
A feature shipped, but the workflow is still slow. Stay with the problem and find out why.
02 / INSIDE AN AI-NATIVE ENGAGEMENT
One customer question.
An entire system.
A retailer wants customers to get useful help with orders and returns. Step through three situations to see the engineering, judgment and teamwork behind the answer.
Support teams jump between order systems, carrier portals and policy documents. Customers repeat themselves. Resolution takes too long.
Which enquiries can we resolve reliably? Map the workflow, examine sample cases and agree a focused pilot with a baseline for handling time, resolution quality and repeat contacts.
CHOOSE A SITUATION
“Where is my order? It was supposed to arrive today.”
A signed-in customer, an order record and a carrier update. The model needs current facts from the business systems.
Your order is in transit. The latest carrier update shows a delay, with delivery estimated for the next working day. That estimate may change; you can check the latest update in tracking.
Authorised order lookup
Latest carrier event
Approved delivery policy
BEHIND THE RESPONSE / READ & EXPLAIN
- 01
Establish whose order this is.
The backend verifies the signed-in customer and their access to the order. It passes only the authorised order context to a narrowly scoped tool. A model-supplied order ID never grants access.
AI-Native Fullstack Engineer - 02
Get facts from the systems of record.
Order and shipment records use consistent identifiers and timestamps. The order-status tool reads live operational data; prepared datasets support analysis and evaluation. Stale or missing events are marked explicitly.
Data Engineer - 03
Ground the response in evidence.
The model uses the approved order lookup and retrieves the relevant delivery policy. It explains the returned facts, links supporting information and separates a carrier estimate from a promise. The team evaluates tool selection and answer quality separately.
AI Engineer - 04
Make the path observable and recoverable.
Traces connect the request, tool call and response. Timeouts and service limits are bounded, and logs protect customer data. Engineers can distinguish a carrier failure from a slow model or application bug.
Cloud Engineer
BEFORE WE SHIP / TEST THE BOUNDARIES
- Another customer’s order is denied.
- The answer matches the latest event.
- An estimate is labelled as an estimate.
AFTER THE PILOT / OUTCOME MANAGER
What changed for the customer?
Track verified resolution and repeat contacts alongside response time. A quick answer is useful when it helps the customer understand what happens next.
Compare similar enquiry types and time periods with the agreed baseline. Explain the limits of the evidence before claiming an impact.
IN DAILY USE / CUSTOMER SUCCESS MANAGER (CSM)
Help the change become a habit.
Step into customer successHelp the retailer’s support champion introduce the workflow and show agents how to check carrier evidence. Review use on eligible order enquiries, ask why some agents still switch tools, and follow up on the specific barrier with the right team. Bring customer feedback into the next success review.
AI IN THE WAY WE BUILD
Use AI.
Keep your judgment.
- 01
Draft a tool contract and test cases with an approved AI assistant.
- 02
Review generated code, permissions and edge cases with the team.
- 03
Run the tests and model evaluations. Keep the evidence that earns a release.
Interactive project simulation with illustrative data and responses. No live customer systems or AI services are connected. Actual assignments depend on your readiness and the project.
Find your part in the work03 / FIND YOUR START
Seven roles.
A world to rethink.
Choose the kind of problem that makes you lean forward. Bring what you’ve learned. Discover what you can build.
AI Engineer
Applied AI and evaluation
Solutions Consultant
Business discovery and AI solution design
Cloud Engineer
Cloud platforms and reliability
Data Engineer
Data foundations for AI and business decisions
Outcome Manager
Customer outcomes, delivery and adoption
AI-Native Fullstack Engineer
AI-native product and software engineering
Customer Success Manager (CSM)
Customer relationships, onboarding and lasting adoption
04 / YOUR NEXT STEP
Your next step.
Starts here.
Apply with your academic details, role preferences and résumé. Already applied? Sign in to take your assessment when your sitting opens.