Systems Engineer

AI Evaluation Engineer, Bespoke Labs

• Contributed task authorship to OpenThoughts, Bespoke Labs' open reasoning dataset initiative.
• Designed Harbor-format Terminal Bench style tasks covering concurrency bugs, semaphore leaks, and write-ordering failures.
• Applied Bespoke-MiniCheck for claim-level factual grounding using NLI-based verification.

Builder and AI Systems Engineer, Promco

• Building an order-to-approval system for liquor distributors that turns WhatsApp orders into commercially reviewable drafts.
• Designed explicit state machines and durable session handling for compliance-sensitive order flows.
• Built typed tool interfaces, versioned prompts, validation, and failure tracing around irreversible business actions.

AI Training and Model Evaluation Engineer, AfterQuery

• Conducted frontier model evaluations for logic correctness, security, and infrastructure workflows.
• Built Project Anvil evaluation environments using base snapshots, gold patches, and automated oracles.
• Created Terminal Bench v2 tasks for concurrency bugs, security vulnerabilities, and infra debugging.

Backend Engineer, Alcora

• Architected a multi-tenant headless ERP with real-time stock and demand intelligence.
• Built payment rails with automated debt reconciliation across brands, distributors, and retailers.
• Reduced promotional wastage by 40% with promo cost versus incremental gross profit tracking.

Founding Engineer, AfyaTelemed

• Architected a telemedicine platform across triage, consultations, pharmacy, and lab integration.
• Built secure video consultation with WebRTC and payment processing across local and international gateways.

Software Engineer, McSystems & Medical Inc.

• Architected a CRM with voice-to-CRM logging, reducing manual data entry by 70%.
• Rebuilt a legacy PHP medical sales marketplace into an offline-first Next.js platform.

CareWeave : AI-Native Healthcare Coordination Platform

An AI-native platform that transforms fragmented, unstructured healthcare documents (images, referrals, scanned PDFs) into a grounded knowledge graph using vector embeddings and relationship-native reasoning. Automates provider-payer loops with low-latency voice agents and human-in-the-loop validation, building institutional memory that reduces referral delays and claim denials.

Taji : Real-Time AI Product Demo Agent

A 24/7 AI agent that joins video calls to deliver multilingual, interactive product demos. Drops into any SaaS product to navigate it via VLM-based agentic automation, qualifying leads and booking follow-ups directly in your CRM and calendar. Built on LiveKit, Deepgram, and Cartesia (<700ms perceived latency).

CVCS : AI Cognition Version Control System

Version control for enterprise AI cognition and context. CVCS tracks decisions, corrections, and contexts to make agent behavior explicit, versioned, and portable—retaining organizational knowledge while keeping agent runtimes interchangeable.