Lunar has an open Forward Deployed Engineer Intern role based in Gurgaon, India. Its official careers page advertises a three-month, full-time summer internship, with a monthly stipend of INR 1,00,000 to INR 2,00,000. The work is in person, including deployment at a manufacturing customer's site; this is explicitly not a remote opportunity.
Jobsii checked the vacancy, careers index and application form on 25 October 2026. The employer is reviewing applications on a rolling basis. Although it calls the programme a summer internship, it does not identify a calendar year, exact start date or closing date. Confirm the intended cycle rather than assuming immediate joining or a specific 2027 intake.
Applicants should be able to demonstrate something they have already built and shipped. Lunar seeks depth in at least one of backend systems, data pipelines or applied machine learning, not necessarily all three. It emphasises practical capability rather than degree or institution, together with problem-solving, ownership, and communication with both engineers and factory personnel.
The internship involves deploying AI agents, connecting them to existing plant systems, building ingestion and retrieval integrations, and developing evaluation code. Conversations with operators inform requirements and product feedback to the founders. No numerical experience minimum, marks cutoff, graduation-batch restriction or vacancy count is published.
Treat the published stipend as a range, not a promise that every successful applicant receives INR 2,00,000. The page does not explain how an individual amount is determined, payment timing, deductions or relocation reimbursement. Ask for those terms in writing. Its full-time offer statement is conditional on exceptional work; an internship offer is not an unconditional permanent-job guarantee.
Before applying, consider whether you can commit to three months in Gurgaon and customer-site work. The form requires confirmation that you can work there in person throughout the internship. General interest in remote AI work is therefore not enough. Travel between locations, site access, working hours and accommodation arrangements need clarification directly with Lunar.
Use the official application page linked below. Full name, email, resume upload and the in-person availability checkbox are required. Phone, GitHub, LinkedIn and portfolio fields are labelled optional. The page asks for a PDF or Word resume up to 4MB. The vacancy also offers careers@trylunar.ai as an email route. Jobsii did not submit an application.
Prepare accessible links to work you can explain, and follow the employer's current instructions if the form changes. No interview sequence, assessment format or selection timetable is disclosed. Keep any confirmation the employer supplies; opening the Jobsii article does not register an application with Lunar.
The remaining sections are independent preparation advice, not extra Lunar requirements or an employer-issued exercise. This role combines engineering and customer understanding. A project explanation should therefore show both what you built and whose problem it addressed, rather than only listing frameworks or model names.
Choose one project where your contribution is clear. Explain the incoming data, the user-facing outcome, a failure you investigated and a tradeoff you made. If the project was a prototype, say so. Do not call it production software merely because it was deployed to a public URL, and do not take sole credit for a team's work.
Consider a synthetic demonstration that answers questions about maintenance records. Use invented records rather than customer or employer data. Each record might contain an equipment identifier, an event time, an issue description and a resolution note. This example concerns information retrieval only, not instructions that control machinery or authorise maintenance.
Start by defining one user question: which recorded incidents mention a recurring issue on a particular asset? Build a small ingestion flow that checks identifiers, preserves source references and reports malformed rows. Decide how to handle a repeated file import so the same record does not silently appear twice.
Next, distinguish event time from import time. A late-arriving record may describe an earlier incident. A useful explanation would show which timestamp your answer uses and why. Similarly, decide how a corrected record replaces or supplements an earlier version. Document that choice instead of leaving two contradictory entries without context.
For retrieval, retain the link between an answer and the records supporting it. Include a question for which no record exists and demonstrate an explicit insufficient-evidence response. If a language model is used, distinguish generated wording from verified record contents. Do not present a plausible answer as evidence that the underlying information is correct.
Create a short evaluation sheet before demonstrating the system. Include the question, relevant record identifiers, expected factual content, observed output and a judgment with reasons. Add duplicate records, missing equipment identifiers and contradictory notes as separate cases. Explain what each case tests rather than reporting one unexplained accuracy percentage.
Finish with boundaries: a synthetic demonstration does not establish readiness for a live factory. Real deployments need employer-approved access, security review, operational supervision and validation appropriate to their use. Showing that you recognise those boundaries is more useful than overstating a small project's reliability or connecting it to real industrial systems without permission.
For the communication side, write a brief handover for a non-technical operator: what information the demonstration uses, what it can answer, what it cannot answer and how to flag an incorrect result. Keep that separate from engineering setup instructions. This makes your project easier to evaluate without assuming every stakeholder understands the code or model.
Ask who supervises the internship, how customer feedback is prioritised and what decisions require approval. Clarify the expected location pattern, any travel and the actual programme dates. These questions help distinguish a workable arrangement from assumptions based only on the attractive stipend headline.
Confirm the selected stipend, payment schedule and any conversion criteria in the written offer. Revisit the official vacancy before submitting because rolling review is not a guaranteed availability period. Jobsii curates the opportunity and provides independent guidance; Lunar handles eligibility decisions, selection and employment terms.
Review the role details, deadline, and work mode before applying. Jobsii tracks the apply action so candidates can keep application history inside their dashboard.
