AI Engineer | Google Developer Expert (ML, Cloud, Kaggle) | Hugging Face 🤗 Fellow
I'm a Machine Learning Engineer currently leading AI development at IntelliTek, where I focus on using Generative AI to enhance the healthcare domain — automating clinical workflows like SOAP note generation, extracting structured data from unstructured conversations, and ensuring HIPAA-compliant ML pipelines.
As a triple Google Developer Expert in Machine Learning, Cloud, and Kaggle, I bring both depth and breadth to real-world AI systems. My work sits at the intersection of NLP, Speech Technologies, and Medical AI.
👉 Consulting? Here’s how I work 🛠️ My GSoC project at TensorFlow
<div class="career-item">
<h4>IntelliTek Products Pvt. Ltd. — <em>ML Engineer</em></h4>
<p class="meta">📍 Sep 2024 – Present</p>
<ul>
<li>Designing GenAI solutions for clinical settings: automated SOAP notes, issue lists, and medical summaries.</li>
<li>Built adversarial-robust, compliant pipelines with data security guardrails under HIPAA.</li>
</ul>
</div>
<div class="career-item">
<h4>TensorLake Inc. — <em>ML Engineer</em></h4>
<p class="meta">📍 Apr 2024 – Aug 2024</p>
<ul>
<li>Designed <strong>Indexify</strong>, a real-time multimodal unstructured data engine.</li>
<li>Improved processing speed by 30% and boosted query efficiency by 25%.</li>
<li><a href="https://github.com/tensorlakeai/indexify">Project Link: Indexify</a></li>
</ul>
</div>
<div class="career-item">
<h4>PrediQt Business Solutions Pvt. Ltd. — <em>Senior AI/ML Engineer</em></h4>
<p class="meta">📍 Jun 2023 – Nov 2023</p>
<ul>
<li>Led LLM finetuning and model optimization efforts.</li>
<li>Achieved 40% faster inference and 50% uplift in model serving throughput.</li>
</ul>
</div>
<div class="career-item">
<h4>Dynopii Inc. — <em>ML Engineer</em></h4>
<p class="meta">📍 Apr 2021 – Mar 2024</p>
<ul>
<li>Engineered full-stack speech and audio ML pipelines for Conversational AI.</li>
<li>Drove 35% increase in engagement and cut training costs by half.</li>
</ul>
</div>
<div class="career-item">
<h4>Celebal Technologies Pvt. Ltd. — <em>Data Scientist</em></h4>
<p class="meta">📍 Sep 2021 – Dec 2021</p>
<ul>
<li>Applied Classical ML, NLP, and CV techniques to enterprise use cases.</li>
<li>Improved algorithm speed by 25% through Python-SQL optimization.</li>
</ul>
</div>
<div class="career-item">
<h4>TensorFlow (Google) — <em>GSoC 2022 Contributor</em></h4>
<p class="meta">📍 May 2022 – Sep 2022</p>
<ul>
<li>Contributed to <a href="https://summerofcode.withgoogle.com/programs/2022/projects/RmEpoyDX">TensorFlow Decision Forests</a>, improving ease-of-use for Kaggle competitions.</li>
<li>Created real-world examples for YDF and interpretability tools, boosting adoption by 15%.</li>
</ul>
</div>
FireRequests is a high-performance, asynchronous HTTP client designed for large-scale ML workloads. It’s used by companies like Roboflow and supports concurrent interaction with providers like OpenAI and Google.
<h4>Key Highlights:</h4>
<ul>
<li>⚡ 10x faster I/O for uploading/downloading large payloads</li>
<li>🔄 Robust retry logic, fault tolerance, and concurrency using <code>asyncio</code>, <code>aiohttp</code>, and <code>aiofiles</code></li>
<li>📦 Seamless support for Jupyter, batch processing, and streaming APIs</li>
<li>🧵 Integrated with <code>nest_asyncio</code>, <code>Semaphore</code>, and exponential backoff for reliability under load</li>
</ul>
- TensorFlow
- PyTorch
- Transformers
- Scikit-Learn
- XGBoost
- LLMs
- Audio ML
- Speech-to-Text
- Text-to-Speech
- vLLM
- GCP
- AWS
- FastAPI
- Flask
- Redis
- TensorRT
- Docker
- PEFT
- TRL
- Optuna
- GitHub Actions
- LangChain
- Weights & Biases
- Kaggle Competitions Master & 2× Expert
- Gold Medalist (Top 10) in RSNA MICCAI Brain Tumor Segmentation Challenge
- Endorsed by experts like Sayak Paul (Hugging Face, GDE)
- Co-Organizer of TensorFlow User Group Kolkata and GDG Cloud Kolkata
- Regular speaker and mentor in the global open-source AI ecosystem
- Hugging Face 🤗 Fellow, building public LLM tools and sharing cutting-edge research
- 📧 Email: heyrishiraj@gmail.com
- 🔗 LinkedIn | GitHub
Building the future of AI — one transformer at a time.