Job Location: The Great Bharat

Job Detail:

About Triomics –

Triomics is developing the technology to ensure that life-saving drugs, vaccines, and medical devices reach the market sooner by streamlining the process of clinical trials. By developing a network of technology-enabled clinical trial sites (hospitals or clinics), we can speed up the clinical trials, accelerate participant recruitment and streamline data collection. Triomics integrates with the different healthcare systems and standardizes that data for the clinical trial workflows in real-time, saving time and money and increasing the quality of the data that is eventually submitted to regulatory bodies. Founded by IIT Roorkee graduates, we are driven by our technology-first approach to solving the inefficiencies in clinical trial data workflows. Triomics is funded by leading institutional investors, namely Y Combinator, Nexus Venture Partners, and General Catalyst.

Building a technology that streamlines data pipelines between 100s of systems and keeps them synchronized with 100% reliability comes with great complexity and many technical challenges. From building a resilient system that can handle all edge-case scenarios without breaking to scaling it in a manner that complies with the different regulatory requirements, working with Triomics provides an intellectually challenging environment setting you up for high growth.

About Role –

Triomics is pioneering a technology that will transform how clinical trial data is managed, and we are looking for an NLP Research Intern to help innovate on one of our specific projects. Working with healthcare data involves dealing with structured medical records to unstructured notes with complex terms, entities, and their relationships. We are building a pipeline to extract useful information from some of such unstructured data sources. While LLMs have advanced significantly over the past two years, their utilization in specific problems like this is still not ready for mass adoption. We are looking for someone who is curious enough to take on the challenge of standardizing the healthcare data of millions of patients and unlocking cures and evidence for some of the most life-threatening health conditions in the world.

The project is in its very early days. You will be involved in designing the problem statement to create a dataset by working with the data collection team of Triomics. You will then build a novel solution from scratch using language models while augmenting them with entity relationships and medical terminologies. The final result should be a state-of-the-art model that is scalable and efficient enough to be deployed in real-world applications and is patentable/publishable.

You should have a solid technical background, experience with language models, proficiency in pytorch, and, ideally, a high level of experience with the deployment of large ML models.

Responsibilities –

1. Working with cross-functional teams to understand the problem deeply so that a contextualized machine-learning framework can be developed
2. Framing the dataset structure and breaking it down into specific use cases, and uncovering the insights in the data to make intuitive decisions for modeling
3. Training, Validating, and Building language models that are tuned to the specific use case and domain while also capitalizing on various publicly available research works
4. Document technical specifications, processes, and results while also closely working with and communicating with a team of various business stakeholders.
5. Building use-and-throw demos to showcase the capability of the developed models so that human feedback can be obtained and analyzed

Requirements –

1. Enrolled in a Bachelor’s, Program in Computer Science, Linguistics or a related field
2. Strong knowledge of NLP concepts, techniques, and algorithms, specifically large language models
3. Proficiency in one or more programming languages, preferably python
4. Experience with deep learning libraries and frameworks such as PyTorch, HuggingFace, Tokenizer
5. Excellent problem-solving and analytical skills
6. Curiosity to work in a domain-specific machine learning environment

Ideal Candidate –

1. Someone who has publications in any of these conferences – EMNLP, NAACL, ICML, ICLR, NeurIPS
2. Someone with a strong experience in working with large language models such as GPT-3, RoBERTa, etc

Why Us?

1. We are revolutionizing a unique industry that has the potential to benefit patients from all over the world.
2. We have had company-sponsored workstations in Bali, Sri Lanka, and Manali and take pride in our purely merit-based, hard-working, yet super fun culture.
3. We are working on a few of the most challenging problems in a highly regulated industry which provide you an opportunity to solve some of the most interesting things

Apply Here

Submit CV To All Data Science Job Consultants Across Bharat For Free