Jiten Oswal
AI Architect. Engg Leader. Founder. Angel Investor.
14+ years building AI & Data platforms at Salesforce, Softbank, Amazon, Matterport, Ava Labs and more. 6 US Patents. AI & System Architect. Engineering leader, published researcher on arXiv. Currently, building at the frontier of AI — infrastructure, products, and the people shaping it.
Impact & Scope
I build AI platforms and the infrastructure that powers them — from Salesforce Einstein's predictive intelligence engine to enterprise AI & Data systems at multiple growth and early-stage startups. 6 US patents in system architecture and cloud infrastructure. Published researcher on arXiv and author of 84+ articles on AI systems. Building the future of AI infrastructure, products and mentoring the next generation of AI builders — always open to founders, engineers, and researchers pushing the boundaries of what's possible.
Patents
06
Companies
10+
Years Exp
14+
Articles
84+
Leadership Roles
Stealth Startup Series A
AI Architect / Product Growth / FDE
AI Platform for Enterprises · Enterprise AI Agents · Open Source
2025–PRESENT
AI Platform & Infrastructure
Sapienly Inc. Exited
Founder / CEO & CTO
AI for Knowledge Workers SMB (Legal-Tech, Healthcare & Voice Agents)
2024–2025
Data & AI Infrastructure
Ava Labs Inc.Top 10 globally by TVL
Engineering Leader / Engineering Manager IV
Data & API Platforms · Infrastructure
2022–2024
Data & System Infrastructure
Matterport Inc. Pre to Post-IPO
Sr. Manager / Staff Engineer
Data & API Platform · System Integration · Data Insights
2021–2022
Enterprise System & Data Platform
Softbank Corp.AI + Data & Deep-Tech
Technical Advisor
Tech Due Diligence · Enterprise SaaS, Data, AI, IoT, Edge Computing, Satellite, Robotics
2020–2021
Growth & Early Stage Fund
Salesforce Inc.
Tech Lead / Lead Engineer
Einstein AI Platform & Apps · Search Infra · Analytics · Sales Apps
2013–2020
Enterprise System & ML Platform
Also at Amazon · ExpressMango (Founding Engineer) · SunGard Global Services · Stanford Executive Program · MS Computer Science, San Jose State University
USC Instructor
AI Agents & AI Copilots
Scholarly Publications
Thought Leadership
84+
published articles on AI Architecture, Infrastructure & Products
Five Primitives for Governing Autonomous AI Agents at Runtime
Governing enterprise AI agents is a runtime problem, not a model-alignment one — five primitives (discovery, identity, governance, attestation, supply chain) for mediating an agent’s action against policy before it takes effect and recording it in a verifiable hash-linked ledger.
arXiv:2608.26696 north_eastSeparating Disclosure from Authorization: Field-Tier Minimization for Agent Action Mediation
Per-field tiering keeps raw agent action parameters out of an append-only ledger while preserving auditability — the ledger commits to a digest of the full parameters, so reclassifying a field changes what is disclosed without invalidating history.
arXiv:2608.25474 north_eastBuilding Bazel Packages for AI/ML: SciPy, PyTorch, and Beyond
Technical talk on open-source AI/ML infrastructure — building reproducible, hermetic build packages for the scientific Python ecosystem.
cfp.pydata.org/seattle2025 north_eastInnovation & Patents
Systems & Methods for Optimal Cost-To-Serve in Public Cloud
Real-time cloud cost optimization framework for application development workflows.
Real-Time Cost-To-Serve Metrics & Cost Attribution in Public Cloud
Mapping cloud spend to individual services and user transactions in real time.
E2E Performance Simulation by Controlling Key Factors of Web Products
Controlling and measuring key factors affecting web product behavior at scale.
Performance Simulation with Cost-Benefit Analysis & Multi-Iteration Support
Cost-benefit modeling across high-concurrency simulation cycles for E2E web products.
Extending Performance Simulation to Any Platform via Metadata
Platform-agnostic simulation through metadata-driven configuration for any cloud or on-premise environment.
Actionable Insights to Optimize E2E Web Product Performance
Automated optimization recommendations generated from end-to-end performance telemetry.
Ecosystem Contributions
-
token
McGinnis Venture Competition
Judge · CMU Swartz Center of Entrepreneurship · 2026
-
token
VC Nest Venu Hackathon
Judge · 2025
-
token
USC Viterbi School of Engineering
Instructor · AI Agents Development · Summer 2025
-
token
AI Hack the Stack
Judge · GMI Cloud, Mountain View · 2025
-
token
Skydeck Labs
Judge · AI Hackathon, Berkeley · 2024
-
token
Cal Hacks
Judge · World's Largest Collegiate Hackathon · 2022
-
token
Idea to Impact Startup Event
Judge & Speaker · 2023
-
token
Salesforce Dreamforce Hackathon
Judge · 2013
-
chevron_right
MIT AI Club @ MIT
Member · 2025–present
-
chevron_right
CMU Swartz Center for Entrepreneurship
Advisor & Mentor · 2025–present
-
chevron_right
Stanford Web3 & AI Research Group
Judge, Mentor · 2023–present
-
chevron_right
DraperU Hero Accelerator – Deep Tech
EIR / Mentor · Draper Venture Fund · 2026
-
chevron_right
VC Fellow @ VC Lab
<10% acceptance rate · 2023
-
chevron_right
Pi VC Fellow @ Pi Ventures
30 selected from thousands · 2020
-
chevron_right
Climate Wisdom Fellowship
Invite-only leadership program for climate & impact leaders
-
chevron_right
Delta Analytics Fellow
25–35 selected from thousands · 2017