Two separate funding awards
Funded twice for Parameter Golf model training: once by OpenAI and once by Runway.
Former Facebook Contractor
Conversions API, Facebook Pixel, and custom JavaScript integrations.
Frontier compute
Provisioned multiple rented systems in every GPU class used: B200, B300, H200, and A100.
Build · operate · evaluate
I'm Francis Clase, an AI systems engineer who received two separate Parameter Golf training awards: one from OpenAI and one from Runway. My portfolio spans model training, agent evaluation, developer tools, location intelligence, and production operations.
Francis Clase
AI systems · data · operations
8,000+
search keywords grown
14 → 16
Next.js production range
B200 → B300
frontier GPU training
6
public OpenAI challenge PRs
Portfolio / 01
Production systems, public technical artifacts, and interface case studies across AI, developer infrastructure, location intelligence, and revenue operations.
Flagship / Location intelligence
A shareable location-intelligence product for Mac, Windows, and mobile workflows: custom search, place discovery, address context, local categories, and an interface designed for practical field decisions.
Private product build · interface case study
A public developer-product concept that transforms OpenAPI and JSON schemas into typed SDKs, live documentation, CLIs, MCP servers, and OpenAI-native agent workflows—with reviewable diffs, governed releases, and evaluation traces.
Model craft / Public artifacts
Received two separate Parameter Golf model-training awards—one from OpenAI and one from Runway—with multiple rented systems in each GPU class used: B200, B300, H200, and A100. Six public submissions span adaptive quantization, JEPA, GPTQ, FlashAttention 3, Muon, and legal compression under a strict 16 MB artifact cap.
View public pull requestsRevenue operating layer
RCG
Lead capture → qualification → scheduling → attribution → follow-up
RevOps / CRM / Customer intake
An end-to-end sales and revenue operations stack connecting custom intake, service logic, lead management, attribution, communications, and conversion-focused customer experiences.
Visit the live experienceE-commerce / Paid media / Local intelligence
A production Texas meat-market storefront paired with Google Ads operations, address and neighborhood intelligence, conversion signals, automated campaign controls, and local-market prospecting.
Visit the live storefrontAgent evaluation / 02
A practical workflow for building expert AI tasks: grounded environments, verifiable answers, explicit scoring rubrics, realistic edge cases, and calibrated difficulty.
01
Turn a real operating problem into a bounded task with source systems, constraints, and a defensible target answer.
02
Connect CRM, account data, email, chat, maps, runbooks, dashboards, or APIs into a realistic working context.
03
Write objective scoring criteria, edge cases, escalation rules, and failure conditions that can be checked—not hand-waved.
04
Run, inspect, refine, and re-test until the task is difficult for the right reasons and reliable in production.
Model operations / 03
10B+
Estimated cumulative tokens used
High-volume hands-on use across research, coding, agent orchestration, evaluation, automation, debugging, and production delivery—not occasional prompt testing.
OpenAI Codex Pro
ChatGPT Pro / GPT-5.6
Claude models
Gemini Pro
Open-source LLMs
Capabilities / 04
Technical depth across the entire path from acquisition and intake to automation, intelligence, and evaluation.
Custom Claude agents, agent-to-agent orchestration, agent harnesses, task environments, model evaluation, scoring rubrics, and human review.
Salesforce-certified CRM architecture, lead routing, account reconciliation, fulfillment workflows, dashboards, scheduling, and lifecycle email.
Geospatial interfaces, custom maps, address intelligence, territory analysis, local search, and data-quality operations.
Meta Conversions API, Facebook Pixel, TikTok and paid-media automation, SEO systems, reputation management, and prospecting.
Python-certified development, Next.js 14–16, TypeScript, JavaScript, APIs, e-commerce automation, and iOS/Android products.
Machine-learning training, prediction networks, Alpaca market automation, multimodal systems, and custom AI video with HyperFrames.
Contact / 05
AI evaluation, revenue operations, location intelligence, automation, or a product that has to work across systems—I'm interested in concrete problems with measurable outcomes.