Find open senior machine learning engineering roles at product companies, funded startups and global engineering centres in India. Weight companies doing real modelling work highest: Google, Microsoft, Amazon, NVIDIA, Adobe, Salesforce, Uber, Sarvam AI, Krutrim, Fractal, Zepto, Swiggy, Flipkart, Meesho, CRED, Razorpay, PhonePe, Navi and Juspay. Each finding must be ONE job posting for ONE role, hosted either on the employer's own careers domain or on an applicant tracking system. Accept job-boards.greenhouse.io, boards.greenhouse.io, jobs.lever.co, jobs.ashbyhq.com, myworkdayjobs.com, smartrecruiters.com, apply.workable.com, jobs.jobvite.com, careers.icims.com, darwinbox.in, keka.com, zohorecruit.com, or a company's own /careers/, /jobs/ or /join/ path that carries a role identifier in the URL. MUST-HAVES, every one of them checkable from the posting itself: the role is based in Bengaluru, Hyderabad, Pune, Chennai, Gurugram, Noida or remote-India, or is explicitly remote within it; the title is Senior Machine Learning Engineer, Senior ML Engineer, Senior Applied Scientist, Senior Data Scientist with a stated modelling and deployment remit, or Staff Machine Learning Engineer, OR the requirements ask for 5 or more years with machine learning named explicitly; The posting must still be open. Reject anything showing 'closed', 'filled', 'position no longer available', 'no longer accepting applications', or a deadline that has passed. The posting must have been published or last updated within the last 21 days. If the page carries no date at all, say so in the summary and rank it below an otherwise identical dated posting rather than assuming it is fresh. Say the ML domain and the frameworks the posting names (large language models, computer vision, speech, recommendations, PyTorch, TensorFlow, JAX). The following RAISE the value of a finding and are NOT requirements: a published compensation band, a named tech stack, the team size, the reporting line, and whether the role is onsite, hybrid or remote. Say which of them are present. Never reject a posting for omitting them: most postings omit compensation, and treating that as disqualifying rejects almost everything. DO NOT SEND: entry-level or mid-level roles; PhD-only research positions with no production component; data-engineering, ETL or analytics roles without modelling; or service-company and consultancy postings from Infosys, TCS, Wipro, Cognizant, Accenture, Capgemini, HCL, LTI Mindtree or Tech Mahindra. REJECT these domains outright, they are never the employer: naukri.com, instahyre.com, foundit.in, monsterindia.com, hirist.tech, cutshort.io, glassdoor, indeed, ziprecruiter, simplyhired, jooble, talent.com, adzuna, reed.co.uk, totaljobs, cv-library, stepstone, xing, jobs-radar.com, itreferraljobs.in, educativ.net, digitalxnode.com, and any page whose title reads 'ROLE at COMPANY | SITENAME'. Also skip search and category pages (any URL carrying q=, query=, keyword=, search=, page= or offset=, or a path ending at /jobs, /careers, /vacancies or /openings with no role identifier), recruiter and staffing agency pages, LinkedIn profiles of people who already hold the role, 'we are hiring' social posts, salary guides, interview-prep articles, and course or bootcamp adverts. Lead with the company, the exact title, the ML domain, and the city.