Subhadip Mitra
Applied AI & Data Executive · Engineering Leader · AI Researcher
I build AI systems that move from research to production at enterprise scale. 15+ years across APAC, currently Head of Data & Analytics for Google Cloud Southeast Asia and Site Lead for the regional professional services organization, and still writing the systems underneath: multi-agent platforms, evaluation infrastructure and GPU inference kernels.
01 Experience
Progressed from Senior Consultant to Head of D&A (Year 2) to Site Lead for all SEA practices (Year 3).
- Practice and site leadership. Built the Data & Analytics practice from zero to eight-figure cumulative delivery value across six countries; as Site Lead, provide delivery governance across the region's AI/ML, Data, Infrastructure, Application Modernization and Security practices.
- Pursuit portfolio. Led and shaped a strategic Data & AI pursuit portfolio across JAPAC, contributing to major competitive services wins, including platform takeouts involving AWS, Azure, Databricks and Snowflake, and supporting substantially larger downstream cloud-consumption opportunities.
- Agent Program Office. Conceived, secured executive sponsorship for and launched a JAPAC Agent Program Office for a 700+ person organization: defined the pilot, governance and efficiency-measurement model and personally built the Agent SDK and automated telemetry backbone for usage, attribution and benefits tracking. Positioned for potential expansion to a 2,000+ person global organization.
- Migration accelerator. Architected and built a four-agent migration accelerator for a leading Asian bank across ~30K notebooks / ~4K ML projects, reducing modeled delivery effort by 97.5% (160K to 4K hours); a representative project moved from about a week to under 30 minutes.
- Platform consolidation. Thousands of models, more than 25 dependency layers, 25 business units: a consolidation treated as intractable by hand. Built a 20+ agent system that reconstructs the lineage graph of a dbt/BigQuery estate, breaks circular dependencies, searches functional, horizontal and vertical consolidations, and then generates and validates every restructured model itself. Rebuilt the most complex business unit in six hours. Turned an effectively unusable platform into a manageable one and became the technical core of a strategic account recovery.
- Account turnaround. Selected by VP leadership for an eight-week onsite intervention to recover a deeply red financial-services migration where the customer was considering abandoning GCP; helped steer the program back to green.
- Executive advisory. Advise CIOs, CTOs, CDOs, CAIOs and CFOs and present in Google Executive Briefing Centers.
Led enterprise AI and data platform transformation for retail banking.
- Partnered with the Retail Bank CIO on data & AI strategy and investment decisions: business-case justification, prioritization and multi-year planning
- Built a data & analytics platform serving 11 markets, 100+ systems and 1,200+ users
- Delivered a self-service ML workbench running 500+ production models, cutting deployment from about six months to about a week
- Built alternative-data credit risk models using news/social signals across 15K+ entities
Architected enterprise data solutions for large enterprises across APAC.
- Designed 5 data lakes with ETL pipelines running up to 1.2 PB/hour and 40K daily files
- Engineered a real-time platform processing 2.5M events/second
- Built ML fraud detection reducing false positives by ~60%
B2B commercial-vehicle marketplace across 15 cities with 25+ OEM and bank partnerships; later led payments-platform and bank-integration engineering in Southeast Asia.
Windows kernel components for Windows 7/8 and Server 2012 R2; early Azure ML and CDN architecture optimization.
02 Research & open source
Quality-diversity evolution (MAP-Elites) for automatically discovering diverse LLM safety vulnerabilities. ICLR 2026 Workshop (AIWILD); extended cross-generational study on arXiv.
Custom Triton kernels: RMSNorm 8.1x faster at 88% of peak memory bandwidth on A100, and a portable W4A16 4-bit GEMM that beats cuBLAS FP16 by 1.1-1.3x in decode workloads, cross-vendor on A100 and MI300X. Speculative decoding at 2-3x in its benchmark setting. Two kernels published to the Hugging Face Kernel Hub.
Sandbagging and deception detection in LLMs via activation probes: 90-96% accuracy in sandbagging and deception-detection settings. Published on PyPI.
Runtime control of LLM agent behaviors via activation steering, without retraining.
Neuro-symbolic framework for automated distributed-system generation. 274x speedup and 60% latency reduction in its benchmark setting. Google Technical Disclosure.
Automated cloud workload discovery system for architecture assessment, built on billion-node graph models with spatiotemporal graph optimization.
03 Selected publications
Reviewer for NeurIPS and ICLR.
All publications & disclosures →04 Selected projects
Industry-agnostic agentic AI for enterprise trust decisions: a tiered cascade of deterministic rules, ML and adversarial multi-agent reasoning with cost-aware routing. JAPAC Champion and Global #2, Google Cloud Professional Services Hackathon 2025, with two associated technical disclosures.
Privacy-preserving consent protocol for LLM training data. Decentralized registry with cryptographic verification and real-time opt-out enforcement.
Java 21 SMPP protocol implementation with virtual threads. 1.8M PDU decodes/sec, 1.5M encodes/sec. Published on Maven Central.