Jinglin Li

🚀 Excited to share our new JAMIA paper! We show how genomic data normalization can run seamlessly inside FHIR Genomics APIs—turning messy, inconsistent variant/HLA/annotation data into clean, standardized outputs. 👉 Why exciting? Because it means clinicians and apps can search, match, and act on genetic data without building custom pipelines. And it’s open source + live to try! Read here: 🔗 https://lnkd.in/gqhgArVu #FHIR #Genomics #Interoperability #PrecisionMedicine #OpenSource

13

1 Comment

Why the EB-3, Not the H1B, Is the Right Path for Foreign Staff-Level Registered Nurses Many international nurses dream of building a career in the U.S. The demand is real and hospitals are hiring, but choosing the wrong visa can waste time and money. Here is the truth: ✅ Staff-level RN roles in the U.S. require only an associate degree for entry. ✅ Because of that, these positions do not qualify for the H1B visa, which is reserved for specialty occupations requiring at least a bachelor’s degree. ✅ The correct path is the EB-3 immigrant visa. The EB-3 was created to address the nursing shortage. It leads to a green card, permanent residency for your family, and a stable career in the U.S. It requires patience, but it works. ⚠️ Important note: This information applies to non-U.S. citizens. Canadian and Mexican citizens may qualify for TN status under the USMCA agreement. If you are a Registered Nurse abroad considering relocation, focus your energy where it counts. The EB-3 is the reliable and proven pathway to permanency in the United States.

5

1 Comment

🤖 How SDLC Is Evolving with RAG & Generative AI in Healthcare IT One important shift I’ve started appreciating while building my Healthcare IT / LSH consulting mindset is this: 👉 Generative AI adds value only when it is embedded correctly into the SDLC — especially in regulated healthcare environments. Through my learning in LSH Tech and Digital LSH, I’ve begun connecting how modern SDLC practices align with RAG (Retrieval-Augmented Generation) and GenAI use cases in healthcare and MedTech. 🔹 How I now look at SDLC in the GenAI era: Requirements & Design: defining compliant, explainable AI use cases Development: integrating RAG to ground GenAI outputs in validated clinical, regulatory, or scientific data Testing & Validation: ensuring accuracy, traceability, and GxP-aligned controls Deployment & Monitoring: continuous monitoring for model drift, bias, and data integrity 🔹 Why RAG matters in Healthcare & MedTech: Reduces hallucinations by grounding GenAI in trusted data sources Supports compliance, auditability, and explainability Makes GenAI usable in clinical, regulatory, and real-world evidence workflows Understanding how SDLC, RAG, and GenAI work together has helped me think beyond hype — and focus on responsible, compliant, and scalable AI adoption in healthcare IT. 👋 Open to connecting with professionals working in Healthcare IT, MedTech, AI, and Digital Transformation roles. #HealthcareIT #SDLC #GenerativeAI #RAG #MedTech #DigitalHealth #LSHConsultant #HealthTech

7

Dispelling a few myths about EB-2 NIW: 1. An approved NIW does not grant you work authorization. You will still need a nonimmigrant visa to be able to stay and work in the U.S. (such as H-1B). 2. An approved NIW does not put you in a different EB-2 backlog (EB-2 = EB-2). It does, however, grant you additional flexibility in that you will not need to start a new PERM with a new employer should you switch employers so that you will be ready to file your I-485 when your priority date becomes current. 3. EB-2 NIW is not only for researchers. Industry professionals can get approved for EB-2 NIW as well (yes, even industry professionals with no research background and no publications). - Interested in EB-1A, EB-2 NIW or O-1? Schedule a consultation through our website! (link on my profile) For questions, email info@waypointimmigration.org or message me and I’ll get back to you shortly. You can also request a free EB-1A or EB-2 NIW evaluation through the "Free Assessments" page on our website! Interested in getting specific examples of profile-building opportunities? Check out the practical profile-building opportunities database in the Immigration Jason newsletter: https://lnkd.in/gin7zj-e Disclaimer: opportunities are not legal advice or guarantees and should be used for brainstorming and inspiration to kick start your journey. #eb2niw #eb2 #greencard #visa #uscis

28

2 Comments

Had the opportunity to join an inspiring AITX meetup in Houston, managed by Michael Daigler, where our breakout group focused on Healthcare AI. I wanted to share a few reflections that stood out from our session: 1. Explainability is Non-Negotiable There’s a strong and growing demand for explainable AI/ML models in healthcare. Clinicians and decision-makers consistently emphasize the importance of transparent outputs—not just predictions, but understandable reasoning. Trust is built when models can be interpreted and justified. 2. Equity is a Priority, Not an Afterthought A major concern raised was the lack of demographic diversity in many health AI datasets. If we don’t address underrepresentation, we risk creating tools that serve some groups well and fail others. Equity and fairness must be built into these systems from the ground up. 3. The “Domain Problem” is Real Many participants pointed out the difficulty in identifying relevant, high-impact clinical problems that are truly ready for AI solutions. While more professionals understand the technology, there’s a gap in understanding which healthcare challenges actually need solving—and how. I am particularly passionate about using EHR datasets for quality improvement through process mining. If this resonates with you, let’s connect and brainstorm at the next AITX meetup! #HealthAI #ExplainableAI #EquityInHealthcare #AITX #HoustonAI #ResponsibleAI #AIinMedicine #AICommunity

13

🚀 Hackathon Win (3rd Prize): AI-Assisted Patient Matching in Healthcare 🏆 During an Oracle internal hackathon, we worked on a real healthcare data problem where patient information flows in from multiple sources — hospitals, clinics, and external networks. This often results in duplicate or near-duplicate patient records, especially when identifiers are missing or inconsistent. Traditional MPI systems can flag these cases, but the ambiguous ones still require significant manual review. That’s the problem we worked on during this hackathon — and we were excited to win 3rd prize for our solution. In just two days, we built a foundational, production-minded pipeline designed to assist with these edge cases — not by replacing MPI, but by augmenting it with explainable, human-in-the-loop AI. 🧠 Solution overview We ingest MPI outputs containing suspected duplicate record groups, apply deterministic data wrangling to handle noisy healthcare inputs, and perform multi-signal matching using probabilistic scoring on structured fields along with semantic similarity via pre-trained embeddings. This results in confidence-based recommendations — merge, review, or no match — supported by an LLM-based explainability layer used strictly post-decision to help reviewers understand why records appear similar. The workflow is surfaced through a reviewer-centric UI concept that supports async processing and fast group-level review. ⚙️ Tech & approach Python, pandas, similarity scoring, embeddings for semantic matching, and LLMs for explanation generation — all designed with auditability, configurability, and human oversight in mind. 📚 Key learning In regulated domains like healthcare, building AI is less about maximizing automation and more about earning trust through transparency and explainability. Huge thanks to my teammates Rupal Kesharwani, Subhajit Khan, Mohit Bhargava and Shivam Sharma for the collaboration and great discussions. A very rewarding hackathon experience. #Hackathon #SoftwareEngineering #AI #MachineLearning #LLM #HealthcareIT #ExplainableAI #Oracle

31

5 Comments

DOCUMENT INCLUDES "HOW TO SEND COMMENTS". "DHS also invites comments that relate to the economic, environmental, or federalism effects that might result from this proposed rule. Comments must be submitted in English, or an English translation must be provided. Comments that will provide the most assistance to USCIS in implementing these changes will reference a specific portion of the proposed rule, explain the reason for any recommended change, and include data, information, or authority that support such recommended change. Comments submitted in a manner other than via http://www.regulations.gov, including emails or letters sent to DHS or USCIS officials, will not be considered comments on the proposed rule and may not receive a response from DHS. Instructions: If you submit a comment, you must include the agency name (U.S. Citizenship and Immigration Services) and the DHS Docket No. USCIS-2025-0370 for this rulemaking. Regardless of the method used for submitting comments or material, all submissions will be posted, without change, to the Federal eRulemaking Portal at http://www.regulations.gov, and will include any personal information you provide. Therefore, submitting this information makes it public. 

1

Are we striking the right balance between profit and discovery? In our Bharat, there's increasing pressure for every research pursuit to show immediate commercial value. While profitability is essential, I wonder if we're leaving enough room for curiosity-driven science. History shows us something interesting: When Plank, Bohr and Einstein first explored quantum mechanics in the early 1900s, they were simply trying to understand matter and energy. That fundamental curiosity eventually gave us lasers, semiconductors, and MRI machines. Michael Faraday's experiments with electromagnetism weren't driven by business plans but pure fascination with invisible forces. Yet his work became the foundation for all electrical engineering. And in life sciences, when researchers first studied mRNA in the 1960s, nobody imagined vaccines. Scientists like Katalin Karikó spent decades on this "impractical" research, often struggling for funding. That patience and persistence? It gave us vaccines that saved millions of lives. Even the internet began as ARPANET an academic project for research collaboration, not a commercial venture. A common observation here is that fundamental research creates the foundation for breakthroughs we can't yet imagine. And I believe Bharat can nurture both, rigorous fundamental research and market-driven innovation. It's about creating space where discovery and application can both thrive. #Innovation #Science #Research #India #Bharat #Discovery #Academia #Industry #LongTermVision

14