Company Description Ace Keystone Consultants (AK Consultants) is a deep-tech innovation R&D; firm based in Coimbatore, Tamil Nadu, India. Founded in 2025, the company focuses on empowering organizations and professionals to achieve their goals through cutting-edge research, strategic advisory, and transformative solutions. AK Consultants supports clients across sectors by driving innovation, helping them navigate complex challenges, and unlocking new growth opportunities.

Core service areas include innovation consulting, research and development for emerging technologies, and customized solutions for sustainable growth.

Job DescriptionWe are seeking a technically rigorous Machine Learning Engineer to join our R&D; and product development team, working on signal, sensor and imaging intelligence. You will convert engineering and product requirements into AI/ML system architectures, data specifications, algorithms, flowcharts and production-grade software, with full ownership of the machine learning lifecycle from problem formulation to validated, deployable models.

We are looking for an engineer who can design and build models and algorithms from first principles, with the mathematical and implementation depth to develop solutions that do not depend on third-party AI APIs or frontier models where those are not the right fit, rather than primarily adapting existing online code, repositories or pre-built solutions.

Technology Focus Areas:

This role spans the following core technology areas across our R&D; programmes:

- Time-series and signal analysis of multi-sensor data (motion, pressure, ultrasound), including filtering, spectral and time-frequency analysis, and feature extraction.
- Anomaly detection and predictive modelling on noisy, non-stationary real-world data, from early detection to risk prediction and decision support.
- Computer vision and image analysis (ultrasound, CT, thermal): segmentation, detection, classification and quantitative analysis.
- Multi-sensor data fusion, physics-informed simulation and synthetic data generation where real-world data is scarce.
- Efficient, edge-deployable models engineered for resource-constrained devices under strict latency, memory and power budgets.

Key Responsibilities:

- Translate engineering and product requirements into system architectures, data flows, algorithm specifications and technical flowcharts.
- Independently develop clean, modular and maintainable Python code for AI/ML applications, including custom algorithms and models where off-the-shelf or API-based options are unsuitable.
- Design, train,



validate and optimise machine learning and deep learning models, and build end-to-end pipelines covering preprocessing, feature engineering, training, inference, validation and decision logic.
- Specify the data required to solve a given ML problem (variables, sampling frequency, resolution, volume, diversity, labels, metadata and quality criteria), and work closely with the Data Team to define collection, preparation, labelling and validation protocols.
- Make evidence-based build-vs-buy decisions: third-party AI API, fine-tuning an existing model, or a custom model, based on accuracy, cost, data privacy, latency and offline requirements.
- Design, build and integrate APIs to serve models and connect them with products and external services.
- Optimise models and software for accuracy, throughput, latency, memory footprint and computational efficiency.
- Develop proof-of-concept solutions for open R&D; problems, and benchmark alternative algorithms and architectures through structured experiments.
- Independently test, debug and troubleshoot ML models and software, and document architectures, algorithms, data requirements, model performance and design decisions.

Required Skills & Experience:

- Strong programming skills in Python, with demonstrated ability to develop applications independently, from problem definition to working code.
- Solid algorithmic thinking, mathematical reasoning (linear algebra, probability and statistics, optimisation) and engineering problem-solving skills.
- Experience in Machine Learning and Deep Learning, with a sound understanding of model internals (loss functions, gradients, regularisation, generalisation), deep enough to build or adapt a model rather than rely only on third-party AI APIs.
- Hands-on experience with Scikit-learn, PyTorch and/or TensorFlow, and proficiency with NumPy, Pandas and data preprocessing.
- Strong understanding of APIs: designing, building and consuming REST APIs, including authentication, error handling, rate limits and versioning.
- Ability to design system architectures, flowcharts, algorithms and data-processing pipelines.
- Rigorous approach to model training, validation, benchmarking and performance evaluation, including metrics selection,



cross-validation and error analysis.
- Experience with time-series or signal data (e.g. filtering, FFT/wavelet analysis, sensor noise handling), or with medical or biological image analysis.
- Experience with software debugging, testing, version control and Git.

Beneficial:

- Awareness of sensors, computing platforms and hardware constraints such as CPU/GPU/NPU capability, RAM, storage, bandwidth and inference latency. The role does not involve hardware design or integration.
- Model optimisation techniques such as profiling, quantisation, pruning and knowledge distillation.
- Computer vision, OpenCV, edge AI, MATLAB/Simulink, Docker or cloud deployment.
- Experience with biomedical or other sensitive data, and awareness of handling it responsibly.

What We Are Looking For:

An engineer who can be given a new technical problem and independently work through:

Problem → Data Requirements → System Architecture → Flowchart → Algorithm → Code → ML Model → Validation → Optimisation The candidate should be able to clearly explain to the Data Team what data is required and why. The use of established libraries, frameworks, APIs and documentation is expected. However, candidates should be able to understand, explain and modify the algorithms and code they implement, rather than depending on solutions they cannot independently develop or troubleshoot.

Qualifications:

- B.E./B.Tech/M.E./M.Tech/M.Sc. in Computer Science, Artificial Intelligence, Machine Learning, Data Science, Electronics, ECE, Mathematics, or a related discipline.
- 2 - 5 years of relevant hands-on experience in ML/AI software development is preferred.
- Demonstrable experience in developing at least one ML/AI solution from problem definition through model development and implementation.

Selection Process:

Shortlisted candidates will undertake an in-person practical technical assessment in Coimbatore, covering:

- Problem analysis and system architecture
- Flowchart and algorithm development
- Identification of ML data requirements
- Independent Python coding
- API design and integration
- ML model development, evaluation, debugging and optimisation
- Communication of data requirements to the Data Team

Use of AI tools is acceptable. The assessment will focus on your own programming, algorithmic thinking and problem-solving ability, so you must be able to explain, justify and modify every part of your solution.

A GitHub profile or portfolio is mandatory. Applications without one will not be considered.

📌 Machine Learning Engineer (Coimbatore)
🏢 aK Consultants
📍 Coimbatore