Jamil Yakasai | Robot Operator · Physical AI & Robot Data Collection

Jamil Yakasai

Robot Operator · Physical AI & Robot Data Collection

I'm an applied-AI product builder and former Sales Operations Analyst moving into physical AI. I've built LLM workflows and data pipelines, I understand how training data shapes model behavior, and I bring a rigorous approach to data hygiene, validation, and process discipline. I want to produce consistent, high-quality teleoperation data engineers can trust.

Robot OperatorTeleoperatorRobot Data Collection Operator

Good robot learning starts with good data.

My work in AI systems taught me one thing over and over: a model is only as reliable as the data it learns from. At Zendesk and Novalyte AI I've cleaned, deduplicated, matched, and validated data so the AI workflows downstream behaved predictably.

Robot learning raises the stakes. I'm motivated to be the operator whose demonstrations are consistent, well-documented, and ready to train on, and who reports issues to engineers clearly.

What I bring to an operator role

These come straight from my experience operating software systems, data pipelines, and customer-facing processes.

Robot operations readiness

Operating complex systems through control interfaces and admin consoles
Monitoring telemetry and dashboards
Repeatable, process-driven execution
Cross-functional collaboration with engineers and leadership

Data quality & annotation rigor

Data cleaning, deduplication, matching, and validation
Data normalization pipelines and CRM data hygiene
Structured outputs and JSON schemas
Understanding of how data quality drives AI/LLM model behavior

Operator readiness

Valid driver's license with a clean driving record, suited to on-site customer deployments and AV operator roles
Comfortable with standing and on-your-feet shifts
Comfortable with repetitive, consistency-critical work

Communication & safety-minded process

Process improvement and workflow architecture
Precise reporting to technical and executive stakeholders
Customer-facing professionalism from enterprise sales support and inside sales

Technical tools

PythonSQLJavaScript / TypeScriptGit / GitHubREST APIsWebhooksSupabase / PostgresBigQueryLLM workflowsPrompt engineeringTool callingSalesforceHubSpot

AI systems work behind the skills

These aren't robotics projects. They show the same habits an operator role needs: running complex consoles, watching telemetry, and keeping data clean enough to trust.

RCC

Revenue Command Center

Novalyte AI · internal operator console

I built a single control plane for clinic acquisition, demand intelligence, workforce exchange, revenue telemetry, and platform administration.

Revenue Command Center Call Center view with a call queue, teleprompter script, and live call panel, shown with demo clinic data

Maps toOperating complex systems through consoles · monitoring telemetry and dashboards

SQL

End-to-end data hygiene

Novalyte AI · Zendesk

At Novalyte I owned SQL data hygiene end to end (cleaning, deduplication, matching), keeping the data that feeds AI workflows accurate and consistent. At Zendesk I ran recurring CRM data-hygiene and validation cycles.

Maps toData quality & annotation rigor · consistency-critical work

LLM

AI automation workflows

Novalyte AI

I engineered AI-enabled workflows for lead capture, enrichment, qualification scoring, routing, and CRM writeback, using LLM workflows, structured outputs, and tool calling.

Maps toUnderstanding how data shapes AI model behavior · structured outputs

Geospatialabs

Founder · independent prototype

I built and deployed a preliminary site-intelligence service for California solar and battery storage (BESS) development. It combines grid, interconnection, parcel, permitting, and environmental context into source-backed briefs, with an interactive Mapbox viewer for parcels, hazard zones, and grid tie-ins.

Geospatialabs homepage at geospatialabs.com, showing the hero headline and a sample parcel preview with a map

Maps toReconciling multi-source data · spatial data

Where I've worked

2024 – Present

San Francisco, CA

Founding GTM & AI Product Builder

Novalyte AI

Built the internal operator Revenue Command Center, a unified control plane for clinic acquisition, demand intelligence, workforce exchange, revenue telemetry, and platform administration.
Owned SQL data hygiene end to end (cleaning, deduplication, matching), keeping the data that feeds AI-enabled workflows accurate and consistent.
Engineered AI-enabled automation workflows for lead capture, waterfall enrichment (Clay, Apollo), qualification scoring, routing logic, and CRM writeback, applying LLM workflows, structured outputs, and tool calling.
Conceived, architected, and deployed Novalyte AI end to end as a multi-sided healthtech ecosystem, shipping patient discovery and intake journeys and clinic onboarding tools. Owned requirements, architecture, and production deployment.

May 2023 – Feb 2025

San Francisco, CA

Sales Operations Analyst

Zendesk

Ran recurring CRM data-hygiene and validation cycles where consistency mattered, and led process-improvement initiatives that improved pipeline visibility and enabled reliable sales-performance analysis.
Built executive dashboards that turned sales activity and pipeline data into conversion, forecasting, and performance insights for revenue leadership.
Supported North American enterprise sales teams through Salesforce workflow optimization, pipeline management, forecasting support, and operational reporting.

Mar 2021 – Feb 2023

Oakland, CA

Inside Sales Representative

SolarCity

Handled repeatable inside-sales customer calls for nearly two years, applying the same process consistently on every interaction and building customer-facing professionalism that carries over to on-site deployments.

Education

MaestroA.A.S., Business, Entrepreneurship & AI Innovation · Mar 2026 – Present

Georgia State UniversityBachelor's degree, Psychology