About the role
structured by ORIAI Engineer (Full-Time) at DeepLure. 📍 Remote during probation → Amaravati, Andhra Pradesh after confirmation · 💼 Full-time DeepLure is an applied-AI startup building the world's first industry-defining critical application of AI in healthcare. In our domain, a model that's "mostly right" isn't good enough.
What you will do
- Read research papers and implement them in PyTorch
- Write training loops from scratch
- Evaluate models against benchmarks, and explain why they fail
- Deploy trained models as production-grade Python inference servers
- Learn new AI concepts fast, and apply them in production
What they are looking for
- Proven end-to-end work: you've designed, trained, evaluated and deployed a model yourself
- Strong deep learning fundamentals and fluency in PyTorch and Python
- Comfort reading papers, including the mathematics involved
- AI tools are allowed here, but you must be able to do all of the above without them, while maintaining quality
- You're willing to relocate to Amaravati after probation
Nice to have
- CUDA, ONNX or TensorRT
- Papers, Kaggle results or open-source ML work
- Vision, speech, multimodal or Indic-language models
Before you apply
- Relocate to Amaravati, Andhra Pradesh after confirmation.
Full posting text
AI Engineer (Full-Time) at DeepLure.
📍 Remote during probation → Amaravati, Andhra Pradesh after confirmation · 💼 Full-time
DeepLure is an applied-AI startup building the world's first industry-defining critical application of AI in healthcare. In our domain, a model that's "mostly right" isn't good enough. We're a small, focused team, and everything you build ships.
The role
We're looking for an AI Engineer who can take a model from research paper to production on their own. You'll read the research, implement it, train it, measure it honestly, and deploy it as a service. You'll own the full lifecycle, not one slice of it.
What you'll do
- Read research papers and implement them in PyTorch
- Write training loops from scratch
- Evaluate models against benchmarks, and explain why they fail
- Deploy trained models as production-grade Python inference servers
- Learn new AI concepts fast, and apply them in production
What you need
- Proven end-to-end work: you've designed, trained, evaluated and deployed a model yourself
- Strong deep learning fundamentals and fluency in PyTorch and Python
- Comfort reading papers, including the mathematics involved
- Years of experience don't matter. What you can do does.
No leased competence!
- AI tools are allowed here, but you must be able to do all of the above without them, while maintaining quality
- If your skill disappears when the AI does, this role isn't for you.
Who thrives here
- You build models because you can't stop thinking about them, not for the paycheck
- You're willing to relocate to Amaravati after probation
To be upfront: this is an early-stage startup. The work is intense, hours can be long, and the problems don't wait. If you're here for the work itself, you'll love it. If you're only here for the paycheck, it won't be a good fit for either of us.
Nice to have
- CUDA, ONNX or TensorRT
- Papers, Kaggle results or open-source ML work
- Vision, speech, multimodal or Indic-language models
Apply
👉 https://lnkd.in/g-5YW8-d
Seniority level: Entry level
Employment type: Full-time
Job function: Engineering and Information Technology
Industries: Technology, Information and Internet