1  PROFILE
 2  Senior ML Engineer with 9+ years of experience building and deploying NLP and LLM systems at production scale.
 3  Expert in taking language model research from prototype to low-latency inference.
 4  Core Identity: Applied Researcher & Systems Builder.
 5  
 6  LEADERSHIP & TRACK RECORD:
 7  - Scaled NLP Teams: Managed teams of 6+ engineers across two product verticals.
 8  - Product Delivery: Shipped production LLM-powered products (document understanding, code generation assistant) used in 12 countries.
 9  - Business Impact: Co-founded a language-AI startup reaching $2.1M ARR, 2 filed patents, featured at NeurIPS 2023.
10  - Governance: Open-source maintainer (18k GitHub stars across 3 libraries).
11  - High-Stakes Delivery: Real-time inference systems serving 50M+ API calls/day for a global fintech platform.
12  
13  TECHNICAL STACK (CORE EXPERTISE):
14  - Domain: NLP & LLM Systems (transformers, RAG, fine-tuning, RLHF, information extraction, conversational AI).
15  - Frameworks: PyTorch, HuggingFace Transformers, vLLM, TGI, LangChain.
16  - Engineering: Distributed training (DeepSpeed, FSDP), low-latency inference, Python, Rust (basic).
17  
18  PROBLEM SOLVING STYLE:
19  - Pragmatic & Delivery-focused: Builds for production reliability, not just benchmark leaderboards.
20  - Adaptability: Comfortable in fast-paced environments with ambiguous requirements.
21  
22  NON-CORE AREAS (DOWNGRADE SCORE, BUT DO NOT AUTO-REJECT):
23  - Pure Web/SaaS: Not interested in backend/frontend work without a heavy ML component.
24  
25  RED FLAGS (ONLY IF CENTRAL FOCUS):
26  *Note: These skills are acceptable as peripheral/minor requirements (<20%), but should not be the main daily activity.*
27  - Pure MLOps/DevOps: Avoid roles centered on Kubernetes setup or CI/CD pipelines without ML depth.
28  - No-code AI: Pure prompt engineering or off-the-shelf wrappers without architectural depth.
29  - "Black Box" Roles: Positions where the engineer cannot access model internals or tune inference.
30  
31  TARGET ROLES & LEVEL:
32  - Hands-On Technical Leadership. I operate best as a "Playing Coach"—shaping the NLP/LLM architecture while actively building the core systems.
33  - The Non-Negotiable: The role must allow deep involvement in model development and inference optimization. Not looking for pure people management removed from the code.