Blessing Softtech

Job Title

AI/ML Engineer

Experience

4–5 years building and deploying machine learning or AI systems

Employment Type

Full-time

Location

Pune, Maharashtra, India

About the Role

Blessing Softtech is looking for an AI/ML Engineer to build AI capabilities into real products and run them reliably in production. This is an engineering role: the emphasis is on shipping working systems — training or integrating models, wrapping them in solid services, and keeping them fast, correct, and affordable once they are live.

You will work across classical machine learning and modern generative AI depending on what the problem needs. If your instinct on seeing a model in a notebook is to ask how it will be served, versioned, monitored, and paid for, this role will suit you.

Key Responsibilities

Model Development

- Build, train, and fine-tune machine learning and deep learning models for the problem at hand.
- Implement NLP, computer vision, recommendation, or forecasting solutions as project requirements demand.
- Evaluate whether to train from scratch, fine-tune an existing model, or call a hosted API — and justify the choice on cost, latency, and accuracy.
- Design evaluation harnesses and benchmark against clear baselines before declaring anything ready.

Generative AI & LLM Systems

- Build applications on large language models — retrieval-augmented generation, agents, structured extraction, and summarisation.
- Implement embedding pipelines and vector search using tools such as FAISS, Pinecone, Qdrant, or pgvector.
- Design, test, and version prompts systematically rather than by trial and error.
- Fine-tune or adapt open models where a hosted API is not the right answer for cost, privacy, or accuracy.
- Build guardrails and evaluation for hallucination, safety, and output quality.




- Track and control token cost and latency in production.

Engineering & Deployment

- Wrap models as production services — REST or gRPC APIs, batch jobs, or streaming pipelines.
- Containerise workloads with Docker and deploy to cloud or on-premise infrastructure.
- Build reproducible training and inference pipelines with proper versioning of data, code, and models.
- Optimise inference — quantisation, batching, caching, distillation, and GPU utilisation.
- Write clean, tested, reviewable code and work within normal engineering practices, not just notebooks.

MLOps & Monitoring

- Set up experiment tracking, model registries, and automated retraining where it is warranted.
- Build CI/CD for model deployment with staged rollout and rollback.
- Monitor production models for accuracy degradation, data and concept drift, latency, and cost.
- Instrument logging and alerting so failures surface before clients report them.

Collaboration & Delivery

- Work with product, backend, and frontend engineers to integrate AI features cleanly into applications.
- Translate business requirements into technical approaches, and say plainly when AI is not the right tool.
- Document architecture, model behaviour, assumptions, and known limitations.
- Support client and pre-sales conversations with realistic technical assessments of what is achievable.

Required Qualifications

- 3–5 years building machine learning or AI systems, with work that reached production.




- Solid Python engineering skills — not just scripting, but code others can run and maintain.
- Hands-on experience with PyTorch or TensorFlow, plus scikit-learn, pandas, and NumPy.
- Practical experience building applications on large language models — RAG, embeddings, prompt engineering, or fine-tuning.
- Solid understanding of the ML lifecycle: data preparation, training, evaluation, deployment, and monitoring.
- Experience deploying models as APIs using FastAPI, Flask, or an equivalent framework.
- Working knowledge of Docker and at least one cloud platform (AWS, Azure, or GCP).
- Strong SQL and comfort working with both structured and unstructured data at scale.
- Proficiency with Git and collaborative engineering workflows.
- Ability to read a recent paper or model card and assess whether it is worth implementing.
- Bachelor's or Master's in Computer Science, Engineering, Mathematics, or a related field — or equivalent practical experience.

Preferred / Good to Have

- MLOps tooling experience — MLflow, Kubeflow, Weights & Biases, DVC, or SageMaker Pipelines.
- Kubernetes and scalable serving infrastructure.
- Model optimisation experience — ONNX, TensorRT, quantisation, or knowledge distillation.
- Familiarity with LLM frameworks such as LangChain, LlamaIndex, or the Model Context Protocol.
- Experience running open models locally or self-hosted (Llama, Mistral, or similar).
- Streaming or real-time inference at scale, and experience with Kafka or similar.
- Distributed or multi-GPU training experience.
- Speech, OCR, or document-understanding pipelines.
- Awareness of AI governance, data privacy, and responsible AI practices.

Pay: From ₹30,000.00 per month

Benefits:

- Paid sick time

Work Location: In person

📌 AI/ML Engineer (India)
🏢 Blessing SoftTech
📍 India