Nervesparks
Artificial Intelligence · Machine Learning · Deep Tech
Job Description — Software Engineer (AI / ML)
ROLE
Software Engineer (AI/ML)
DEPARTMENT
Engineering — AI & ML
LEVEL
L1 / L2 (based on exp.)
LOCATION
On-site — Delhi
ABOUT NERVESPARKS
Nervesparks is an AI-first technology company building intelligent systems and agentic frameworks that solve real-world problems at scale. We sit at the frontier of machine learning research and applied product engineering — the place where academic rigour meets production-grade deployment. Our teams are diverse, deeply technical, and driven by genuine intellectual curiosity.
We are looking for engineers who think in models, build with intention, and care about the quality of what they ship. If you have spent your academic and early professional life going deeper on Data and AI/ML — not just using frameworks but understanding what is underneath them — Nervesparks is where that curiosity will be put to serious use.
ROLE OVERVIEW
As a Software Engineer (AI/ML) at Nervesparks, you will be responsible for the full lifecycle of machine learning and data solutions — from framing the problem and selecting the right approach, through model development, experimentation, evaluation, and deployment into production systems. You will work closely with research, product, and infrastructure teams, and your work will have direct, visible impact on the capabilities we deliver to clients and end users.
This is not a role for someone who applies pre-built APIs and calls it machine learning. We expect you to understand your models, defend your design choices, and continuously improve what you build.
WHAT YOU WILL DO
› Design, develop, train, and evaluate machine learning and deep learning models across a range of problem domains — including but not limited to NLP, computer vision, recommendation, and predictive analytics.
› Translate business and product requirements into well-scoped ML problem statements, and lead the end-to-end development from data exploration through model deployment.
› Work with large, complex, and often imperfect datasets — performing data engineering, feature engineering, and exploratory analysis as an integral part of model development.
› Experiment rigorously — design experiments, track results, interpret findings, and communicate what worked, what didn't, and why.
› Deploy models into production pipelines and maintain awareness of model performance over time, including monitoring, drift detection, and retraining.
› Collaborate with peers on architecture decisions, code reviews, and technical design discussions — contributing to a high-quality engineering culture.
› Stay current with developments in AI/ML research and bring relevant ideas from the literature into the team's applied work.
› (Where applicable) Contribute to the organisation's research output — writing internal technical reports, and where work meets the bar, contributing to external publication.
EDUCATIONAL QUALIFICATIONS
Qualification
Notes
Required
B.Tech in Artificial Intelligence & Machine Learning, Computer Science (AI/ML specialisation), or a closely related engineering discipline
Must be from an accredited institution. Coursework should demonstrate a strong foundation in mathematics, probability, linear algebra, and core ML theory.
Preferred
M.Tech in Artificial Intelligence, Machine Learning, Data Science, or a research-oriented postgraduate programme in a related field
Candidates with an M.Tech are particularly valued where their thesis or coursework demonstrates independent research capability and depth of technical contribution.
EXPERIENCE
Required: 0–3 years of professional or project-based experience in machine learning, data science, or a related applied AI field. For exceptional candidates with strong academic and research profiles, we will consider applications with limited or no formal work experience.
› Hands-on experience building, training, and evaluating machine learning models — whether in an industry role, internship, academic project, or personal research project.
› Experience working with Python as the primary language for ML development. Proficiency with core libraries: NumPy, Pandas, Scikit-learn, and at least one deep learning framework (PyTorch strongly preferred; TensorFlow acceptable).
› Familiarity with the ML development workflow: data preprocessing, feature engineering, model selection, hyperparameter tuning, evaluation, and iteration.
› Experience working with real datasets — messy, incomplete, or large-scale data is a plus.
› Exposure to model deployment or productionisation, even at a small scale, is valued.
RESEARCH, PUBLICATIONS & PROJECTS
Nervesparks places significant value on candidates who have engaged with the broader AI/ML research community or have demonstrated initiative through independent technical work. The following are actively sought — and will be weighted meaningfully in the evaluation process:
Area
What We Look For
Why It Matters to Us
Published Research Papers
Peer-reviewed papers at recognised AI/ML conferences (NeurIPS, ICML, ICLR, CVPR, ACL, EMNLP, AAAI, or similar) or in reputable journals. Workshop papers, arXiv preprints, and co-authored student papers are also considered.
Published research demonstrates the ability to formulate a problem rigorously, conduct systematic experiments, and communicate findings at a standard the field accepts. This is a strong differentiator.
B.Tech / M.Tech Thesis
A thesis or capstone project in an AI/ML domain that demonstrates independent research capability — not merely application of existing tools but contribution of original experimentation, analysis, or methodology.
Thesis quality often reveals depth of thinking that coursework grades cannot. We read theses. A strong thesis is weighted equivalently to industry experience at this level.
ML / AI Projects
End-to-end projects — on GitHub, Kaggle, Hugging Face, or equivalent platforms — that demonstrate problem scoping, model development, experimentation, and deployment. Projects should be documented and reproducible.
Projects show initiative, curiosity, and practical capability. We look for projects where the candidate has made genuine choices — not followed a tutorial step-by-step — and can explain and defend every decision.
Kaggle / Competition Rankings
Strong performance in competitive ML challenges (Kaggle, DrivenData, AICrowd, or similar). Top-decile finishes in well-attended competitions are particularly noted.
Competitions test applied ML thinking under constraints — a useful signal of resourcefulness, experimentation discipline, and performance orientation.
TECHNICAL SKILLS
Domain
Skills & Technologies
Languages
Python (primary — strong proficiency required). Working knowledge of SQL. Familiarity with C++ or Julia is a plus.
ML Frameworks
PyTorch (strongly preferred), TensorFlow / Keras. Familiarity with JAX is a plus.
ML Libraries
Scikit-learn, NumPy, Pandas, SciPy, Matplotlib / Seaborn. Familiarity with Hugging Face Transformers, LangChain, or similar LLM tooling is valued.
Model Development
Supervised and unsupervised learning, deep learning (CNNs, RNNs, Transformers), feature engineering, hyperparameter optimisation, cross-validation, model evaluation and interpretation.
Data Engineering
Data cleaning and wrangling, EDA, working with structured and unstructured data. Experience with large-scale data (Spark, Dask, or cloud-based data pipelines).
MLOps / Deployment
Familiarity with experiment tracking (MLflow, W&B), model versioning, containerisation (Docker), and basic REST API development for model serving. CI/CD exposure is a plus.
Mathematics
Strong foundation in linear algebra, probability & statistics, calculus, and optimisation theory. Ability to read and implement algorithms from mathematical notation in research papers.
Version Control
Git — comfortable with branching, pull requests, and collaborative development workflows.
WHAT WE ARE REALLY LOOKING FOR
Beyond the formal requirements, the qualities that define the people who thrive at NerveSparks are:
› Intellectual depth — you go beyond using tools to understanding why they work, when they fail, and what alternatives exist.
› Experimental rigour — you treat model development as an empirical discipline, not a trial-and-error exercise.
› Communication clarity — you can explain your work and your reasoning to both technical and non-technical audiences without oversimplifying.
› Ownership — you take end-to-end accountability for what you build, from data to deployment.
› Curiosity — you follow the literature, explore new techniques, and bring ideas from outside your immediate work.
› Integrity in results — you do not oversell model performance, you flag uncertainty, and you are honest about limitations.
GOOD TO HAVE
› Experience with Large Language Models (LLMs), fine-tuning, prompt engineering, or RAG-based architectures.
› Familiarity with reinforcement learning or multi-agent systems.
› Contributions to open-source AI/ML libraries or datasets.
› Experience with cloud ML platforms (AWS SageMaker, GCP Vertex AI, Azure ML).
› Exposure to edge deployment or model compression (quantisation, pruning, distillation).
› Understanding of responsible AI principles — fairness, explainability, bias detection, and mitigation.
HOW TO APPLY
Send your application to [[email protected]] with the subject line: Software Engineer (AI/ML) — Application. Please include:
› Your CV / resume
› Links to your GitHub, Kaggle profile, Google Scholar, or any relevant portfolio — we read these before we read your CV.
› If you have published papers, include links or attach PDFs.
› A short note (under 200 words) on the ML problem or project you have found most interesting — not what impressed others, what genuinely interested you and why.
Nervesparks is an equal opportunity employer. We evaluate candidates on the depth and quality of their technical thinking — not their pedigree, institution, or background. All qualified applicants are encouraged to apply.
Nervesparks · nervesparks.com · [Delhi, India] · [email protected]
This document is confidential and intended solely for recruitment purposes.
Pay: ₹400,000.00 - ₹500,000.00 per year
Benefits:
- Leave encashment
- Paid sick time
Work Location: In person
