Target Roles
My primary identity is AI Engineer. I am targeting roles where modern AI application engineering, enterprise delivery, and data or risk domain depth reinforce one another.
Primary: AI Application Engineering
My strongest near-term fit: building reliable AI applications that combine LLMs, retrieval, tools, evaluation, data systems, and production integration.
- Built a production fraud-detection agent serving 15M API calls per day under strict data-residency constraints
- Designed RAG, hybrid retrieval, LoRA adaptation, and human-in-the-loop controls for high-risk decisions
- Built an enterprise ChatBI / NL2SQL system with tiered routing and ReAct-style tool orchestration
- Worked across Python, SQL, ETL, XGBoost, FAISS, Hadoop, Spark MLlib, and standardized APIs
- Delivered systems from problem definition through deployment, integration, monitoring, and adoption
Forward-Deployed & Solutions Fit
A strong fit for roles that require engineers to work close to customers, constraints, and adoption—not only model code. I bring solution design, API standardization, secure deployment, enterprise integration, and multi-scenario rollout experience.
- Scaled a regulated AI platform to 25+ enterprise and device-manufacturer clients
- Designed around security, data residency, latency, uptime, and human accountability
- Abstracted complex capabilities behind standardized APIs for repeatable delivery
- Connected technical choices to operational workflows and measurable adoption
- Worked across telecom, e-commerce, fintech, logistics, and public-sector scenarios
Domain-Advantaged AI Roles
My durable advantage is domain depth. I have built credit, fraud, recommendation, user-intelligence, logistics, and enterprise analytics systems where data quality and business risk matter as much as model behavior.
- Built credit intelligence using 1,500 behavioral and device features with XGBoost
- Supported 500M+ scoring calls, 2M+ users, and 40+ financial institutions
- Built recommendation, profiling, blacklist mining, and logistics diagnosis systems
- Understands model quality in the context of approval rates, adoption, complaints, and workflow time
- Can bridge legacy data platforms with modern LLM application stacks
Want to validate fit quickly?
Scan the case studies, then contact Theo for roles that need reliable AI applications, enterprise delivery, and data or risk depth.