AI Intern – Intelligent Home Appliances

Location: Bengaluru (Onsite)

Duration: 6 months | Full‑time

About Mankomb

Mankomb is pioneering “zero‑effort sustainability” through AI‑native home appliances such as Chewie and Chewie‑X—devices that turn everyday kitchen waste into value while fitting seamlessly into modern urban living. Backed by deep‑tech veterans from Bosch, Myelin Foundry, and IIT‑IIM alumni networks, we are on a mission to make Indian engineering the global benchmark for intuitive, planet‑positive hardware.

Why this internship matters

Our next product iterations demand smarter perception, adaptive control, and hybrid intelligence that runs both on‑device and in the cloud. As an AI Intern, you will sit at the crossroads of edge ML, firmware, and user‑facing software—turning raw sensor streams into real‑time decisions that delight users and shrink their environmental footprint.

What you’ll do

Hybrid ML Pipelines: Design models that can execute on an NXP eIQ NPU for low‑latency tasks, with cloud fall‑back for heavier inference; build decision logic to switch seamlessly between the two.

Data Engineering & Annotation: Build scalable pipelines to ingest images, audio, and multi‑modal sensor data from field pilots; design labelling guidelines that balance accuracy with speed.

Model Development: Prototype and train lightweight CV and TinyML models (e.g., vision‑based waste classification, anomaly detection, predictive maintenance) in Python/PyTorch/TensorFlow Lite.

Edge Deployment: Quantize/compile models for NXP NPUs, measure latency/energy trade‑offs, and create reproducible build scripts with Yocto/Docker.

Firmware Collaboration: Pair with the firmware team to integrate ML inference into RTOS tasks, tune I²C/SPI data flows, and expose inference results via protobuf or shared‑memory queues.

Interaction Layer Prototyping: Develop lightweight desktop/mobile dashboards (Flutter/React) and over‑the‑air update utilities so users—and the team—can visualise model outputs and trigger tests.

Evaluation & A/B Testing: Define metrics, set up hardware‑in‑the‑loop rigs, run statistical analyses to compare model versions across edge and cloud paths.

What you’ll learn

• End‑to‑end product AI lifecycle—from raw data to shipped firmware and UI hooks
• Practical constraints of running ML on low‑power embedded devices and the cloud
• Rapid prototyping with 3‑D printed fixtures, sensor rigs, and cloud simulators
• Agile hardware‑software co‑development in a fast‑growing startup

Must‑have qualifications

• Recently completed a Bachelor’s/Master’s in CS, EE, Data Science, or related field
• Solid Python skills; experience with at least one deep‑learning framework
• Familiarity with computer‑vision or signal‑processing fundamentals
• Comfortable writing C/C++ for embedded targets and JavaScript/TypeScript or Dart for simple UX layers
• Hands‑on attitude and the curiosity to debug both code and circuits
• Excellent communication—able to distil complex ideas for non‑AI teammates

Nice‑to‑haves (But, we can always learn this on the job)

• Exposure to NXP eIQ™, TensorFlow Lite Micro, or ONNX Runtime on edge devices
• Experience orchestrating cloud inference pipelines on AWS/GCP (SageMaker, Vertex AI)
• Familiarity with MQTT/REST for IoT messaging and OTA updates
• Knowledge of version‑controlled ML ops (DVC, MLflow)
• Participation in Kaggle, robotics clubs, or open‑source hardware communities

What we offer

• Mentorship: Work directly with our CTO who has shipped edge‑AI products globally.
• Stipend of upto 25,000/- per month
• Impact: Your code will run in real kitchens, reducing methane emissions from Day 1.
• Growth: Possibility of a pre‑placement offer (PPO) for high performers.
• Culture: Flexible hours, open design reviews, monthly demo days, and a maker‑space stocked with sensors, dev boards, and fresh coffee.

How to apply

Apply directly on LinkedIn and

Send the following to [email protected] with subject line “AI Intern – <Your Name>”:

1. Résumé/CV (PDF)
2. Portfolio or GitHub links showcasing relevant projects
3. A short note (≤ 200 words) on one everyday sustainability problem you’d solve with AI and why

Applications are reviewed on a rolling basis. Early applicants have an advantage.

Join us in making sustainable living effortless, intuitive, and fun—one intelligent appliance at a time.