Back to jobs
About this role:
MantriAI is the search partner for this role. All applications come to us first; only shortlisted profiles reach the client.
The hands-on modelling seat. The platform runs numerical ML on data from heavy machines. You build the forecasting, anomaly detection and degradation models on that data and own them through to production.
What you'll do:
- Build and ship models on machine and sensor data that drive direct customer value.
- Work the hard version of the problem - failures are rare, labels are scarce, the modeling has to earn its answers.
- Own model quality and evaluation rigor. Set the code bar across the ML codebase.
Who should apply:
- 3+ years hands-on ML post-Bachelor's (2+ post-Master's, 1+ post-PhD).
- You have modeled time-series, sensor or machine data - industrial, IoT, or adjacent domains like fraud, risk, observability where the modeling ran deep.
- You can describe one row of your data - channels, sampling rate, what failure you were predicting. If that sentence is easy, apply.
- The loop includes an original coding problem, not LeetCode. Most strong candidates find it hard.
- If your last two years are retrieval pipelines, agents or fine-tuning, this is not your seat.
Bonus:
- Predictive maintenance, condition monitoring, RUL or reliability work. Vibration or spectral analysis.
- Publications, patents, competitive programming, quant or markets background.
Read this before applying:
- Location: remote (India) for now. Bengaluru office planned for late 2026 - be open to it.
- Notice: shorter is better
Process:
- Screening call with MantriAI
- Intro call with the CTO - ex-quant, ex-big-tech, hands-on. He takes this call himself.
- Coding round, original problem, designed by him. Then fundamentals, system design, founder round. Expedited for strong profiles.
Posted by MantriAI, search partner on this mandate. Shortlisted candidates hear from us within 2 working days.
Machine Learning Engineer
MantriAIBengaluru, Karnataka, IndiaPosted 1h ago
AI Engineer
Job Description
About this role:
MantriAI is the search partner for this role. All applications come to us first; only shortlisted profiles reach the client.
The client is a well-funded, seed-stage startup building AI for heavy industry - predictive maintenance, diagnostics and root-cause analysis on industrial assets. First engineering team now being built in India. Company name shared after the screening call.
The hands-on modelling seat. The platform runs numerical ML on data from heavy machines. You build the forecasting, anomaly detection and degradation models on that data and own them through to production.
What you'll do:
- Build and ship models on machine and sensor data that drive direct customer value.
- Work the hard version of the problem - failures are rare, labels are scarce, the modeling has to earn its answers.
- Own model quality and evaluation rigor. Set the code bar across the ML codebase.
Who should apply:
- 3+ years hands-on ML post-Bachelor's (2+ post-Master's, 1+ post-PhD).
- You have modeled time-series, sensor or machine data - industrial, IoT, or adjacent domains like fraud, risk, observability where the modeling ran deep.
- You can describe one row of your data - channels, sampling rate, what failure you were predicting. If that sentence is easy, apply.
- The loop includes an original coding problem, not LeetCode. Most strong candidates find it hard.
- If your last two years are retrieval pipelines, agents or fine-tuning, this is not your seat.
Bonus:
- Predictive maintenance, condition monitoring, RUL or reliability work. Vibration or spectral analysis.
- Publications, patents, competitive programming, quant or markets background.
Read this before applying:
- Location: remote (India) for now. Bengaluru office planned for late 2026 - be open to it.
- Notice: shorter is better
Process:
- Screening call with MantriAI
- Intro call with the CTO - ex-quant, ex-big-tech, hands-on. He takes this call himself.
- Coding round, original problem, designed by him. Then fundamentals, system design, founder round. Expedited for strong profiles.
Posted by MantriAI, search partner on this mandate. Shortlisted candidates hear from us within 2 working days.