$200 / month
Start free, decide later
No spots left, but you can save the mentor to your wishlist to get notified about their open spots!
Hi, I’m Shekhar.
I am a Senior Machine Learning Engineer with 10+ years of experience designing and deploying large-scale NLP, Search, and Generative AI systems in production. My work spans end-to-end ownership of Query Understanding, Representation Learning, Retrieval, and LLM/SLM evaluation pipelines that serve millions of requests daily at sub-200 ms latency.
Across companies like DoorDash, UnitedHealth Group, and Highmark, I have led initiatives that improved retrieval recall from ~79% to 95%, reduced claim and case resolution times by up to 85%, and built robust LLM evaluation systems (“LLM Judges”) that assess models across 15+ qualitative, compliance, and safety dimensions. I specialize in fine-tuning embedding models, SLMs, and LLMs (LoRA, PEFT, DDP), architecting hybrid search systems (BM25 + vectors), and operationalizing agentic workflows for data annotation, evaluation, and RAG.
Who I mentor and how I help:
I work best with engineers and researchers who want to move beyond theory and build real, production-grade ML systems. Typical mentees include:
• ML Engineers aiming to break into Search, Retrieval, or Relevance teams
• Engineers transitioning into Generative AI, RAG, or Agentic AI roles
My mentoring style is highly applied and system-oriented. Together, we can:
• Design and fine-tune embedding models and small language models for real-world retrieval and routing tasks
• Build end-to-end RAG systems with proper evaluation, monitoring, and cost controls
• Develop LLM evaluation frameworks (LLM-as-Judge, error taxonomies, offline + online metrics)
• Strengthen ML system design thinking, from data pipelines to inference optimization and deployment
You can expect concrete guidance, architecture reviews, hands-on debugging, and honest feedback grounded in production experience.
If you are serious about building scalable NLP and Generative AI systems and want mentorship that emphasizes depth, rigor, and real-world impact I’d be happy to work with you. Book a session, share your goals, and we’ll create a focused plan to help you grow faster and with confidence.
You tell Shekhar your goals. They see exactly what you want to work on before session one.
You meet within days. Shekhar reviews your application and books your first session, usually within a day or two.
You leave with a plan. Concrete next steps and a weekly rhythm from your very first week.
Shekhar Tanwar
Machine Learning Engineer at United Health Care
Enter your email and we'll let you know when Shekhar has open spots. You'll also get a link to manage your wishlist – no account needed.
Book a free intro call with Shekhar
Connect with Shekhar in a quick call (usually 15-20 minutes)
What's included in the trial?
Every trial is a little different–here's what Shekhar says about their trial:
We'll meet for a quick call (usually under 30 minutes) where we:
- Break the ice and introduce ourselves
- Discuss your goals and how I can support you
- Coordinate logistics like time zones, meeting frequency, and communication methods
