BayOne Solutions
Website:
bayone.com
Job details:
Job Title: AI/ML Engineer
Location: Bengaluru, KA (Hybrid - 3 Days/Week onsite)
Job Type: Fulltime
Role Definition
- Looking a genuine AI data scientist who can understand and use models based on context — not someone narrowly limited to a single tool or technique.
- Python expertise is non-negotiable — the candidate must know Python in and out; no learning on the job.
- Candidate should have breadth across the AI ecosystem: classical ML (e.g., XGBoost, SVM — data cleaning, feature engineering, training, cross-validation), NLP techniques (embeddings, text classification), and effective LLM usage (prompt engineering, training/fine-tuning small models/SLMs on corpus data).
- MCP was clarified as just one part of the broader context — does not want candidates restricted to it; they must be adaptable to pivot to NLP, model retraining, etc., without fumbling on fundamentals.
2. Production-Grade Mindset (Key Theme)
- Solutions are all production-grade and customer-facing — he stressed the significant gap between POC and production (scale and complexity are far higher).
- Candidates must come in understanding they are building customer-facing, production-grade solutions.
4. Ideal Background / Domain Fit
- Strong preference for candidates from product-based companies, ideally from the services org of a product company (e.g., own background at Philips Research servicing devices).
- Should understand concepts like log files and customer cases — directly relevant to the DEA/NSP customer-services domain.
- Must have worked with unstructured data (text, voice) rather than purely structured/numeric data (e.g., banking/financial candidates would struggle with the domain context and data type).
5. Top 5 Must-Have Skills
- Hands-on Python development (expert level; not team management)
- Agentic AI (confirmed as more important than GenAI)
- GenAI
- Understanding of the whole AI ecosystem / models
- Product-based service org understanding (how a service org works)
Click on Apply to know more.
_1548133325694.png){kind=logo}