What Building an AI Tool for Regulated Teams Taught Me About Trust
I spent the past year building an AI tool for compliance and risk teams — largely solo, from the AWS infrastructure to the retrieval pipeline.
I ship complex technology into the real world.
I'm Waheed Zarif — a PMP-certified engineer-operator with 14 years turning hardware-meets-software products into deployed, adopted, dependable reality. I work where most people don't: the seam between the physical (IoT, sensors, building systems), the digital (cloud, SaaS, and hands-on AI), and the customer.

My career started at the intersection of hard science and commercial reality. At Celdara Medical, a biotech incubator, I learned to translate complex, high-stakes concepts — CAR-T immunotherapy, CRISPR/Cas9 gene editing — into viable products inside a tightly regulated R&D environment, including authoring the risk framework for a $490M partnership.
That discipline became the foundation for the rest of my career: taking complex, integrated systems from concept all the way to the field. Over a decade at Delos Living, I did exactly that for a portfolio of wellness IoT and PropTech products — designing sensor systems, commissioning them across thousands of real installations, enabling a national dealer channel, and ultimately guiding a product line through a disciplined end-of-life transition. I've deployed technology everywhere from luxury homes to 100 public school districts to US military housing — usually where failure isn't an option.
More recently, I went hands-on with the modern stack. I built and launched Egret AI — a RAG-powered compliance SaaS — entirely myself: AWS infrastructure as code, a full-stack application, and integrations across 20+ language-model APIs. I did it to stay genuinely fluent in the technology I lead, not just fluent about it.
What I do best: make the hard, last-mile part of technology actually work — the deployment, the integration, the customer adoption, the enablement that turns a product into something people rely on and pay for. I'm equally at home owning a field rollout, running a technical sale, or whiteboarding a system architecture.
I'm drawn to IoT and PropTech, applied AI (especially for regulated industries), and mission-critical and defense technology — domains where reliability, rigor, and the ability to bridge the digital and physical actually matter.
My core expertise lies in:
Vice President — Wellness Technology, Deployment & Lifecycle
Delos Living LLC, Los Angeles, CA
2020 - Present
I lead the technical last mile across Delos's IoT/PropTech portfolio. I directed the deployment of 2,400+ networked sensors across 100 school districts and on-site programs for US military housing, served as technical-sales lead for a 70+ dealer channel, and executed a product line's end-of-life transition with zero critical customer escalations. I helped manage a ~$4M program portfolio and coordinated distributed engineering teams across the US and India.
Senior Director & Director — Program Management & Deployment
Delos Living LLC, New York, NY
2015 - 2020
I progressed through roles of increasing scope: driving development of the DARWIN™ integrated hardware/software platform, building a global hardware supply chain (China, Canada, Netherlands), commissioning the platform across 40+ properties with Crestron and Lutron integrations, and mentoring a team that grew from 3 to 10.
Founder & Lead Technical Architect
Egret AI, Los Angeles, CA
2025 - Present
I solo-built and launched a B2B SaaS platform delivering RAG-powered compliance advisory for regulated industries — architecting a secure multi-tenant AWS environment via Terraform, a Next.js + Django full stack, and a retrieval pipeline integrating 20+ LLM APIs with zero model-training data exposure.
Co-Founder & Principal
Nexis Management, Los Angeles, CA
2024 - Present
I provide product-discovery and program-architecture services — including leading 0-to-1 discovery for an international client's enterprise SaaS platform, authoring the PRD and navigating cross-border data-sovereignty compliance.
Research Scientist
Celdara Medical, Lebanon, NH
2012 - 2015
I conducted hands-on R&D in a highly regulated biotech environment — advancing CAR-T and CRISPR/Cas9 therapeutic work, authoring the risk-monitoring framework for a $490M partnership, and running due diligence on 20+ early-stage assets.
Bachelor of Engineering, Biotech Engineering
Thayer School of Engineering at Dartmouth College
2011 - 2012
Bachelor of Arts, Engineering Sciences
Dartmouth College
2007 - 2011
International Baccalaureate
UWC - USA
2005 - 2007
Certifications
PMP (Project Management Institute, 2024)
DevOps Deployment Automation: Terraform/AWS/Docker (2022)
Python for Data Science (2019)
WELL AP (2017)
Lean Six Sigma Green Belt (2012)
Large-scale, regulated deployments from design through commissioning, integration, and adoption.
AWS, Terraform, full-stack (Next.js/Django), applied AI (RAG/LLMs), IoT & systems integration.
Turning complex technology into value for customers, channel partners, and executives.
Working across product, engineering, operations, and deployment teams.
PMP-certified; complex, cross-border, full-lifecycle delivery with a risk-first mindset.
Cultivating and negotiating with partners, vendors, and corporate clients
I write about the things I work in and think about — applied AI and machine learning, program and risk management, and the occasional bit of math I find beautiful. It's where I keep my hands (and head) in the technical details.
I spent the past year building an AI tool for compliance and risk teams — largely solo, from the AWS infrastructure to the retrieval pipeline.
In the age of AI and geopolitical instability (tariffs), traditional risk management needs to evolve and adopt agile processes.
A series of short articles on image recognition using Keras API
A series of short articles on image recognition using Keras API