Sebastian Becerra

Sebastian
Becerra

I build the systems I run on.

Get in touch
Based in
Miami, FL
Focus
AI systems + infrastructure
Experience
9+ years
Languages
EN · ES · PT
01The short version

I'm a software engineer who spends his days inside the operations he builds for. Nine years across logistics networks, freight marketplaces, and the planning systems behind them, writing the TypeScript in the morning and living with the consequences on the dispatch floor in the afternoon.

That loop is the whole point. I design internal systems of record, data models, and AI-assisted automation, then I actually run on them. It tends to produce software that survives contact with a real operation, because the person who shipped it is also the person the pager wakes up.

Measured in production
98%

On-time against contract SLAs

3,000+

Trucks of capacity planned

300+

Carrier companies onboarded

90+

Deliveries planned per day

97%+

Operational data accuracy

35%

Fewer inter-facility delays

18%

Throughput gain from workflow redesign

9+

Years running live operations

02selected work

Selected work

Two systems that ran in production, with the numbers they actually produced.

01Marketplace · AI

Ship AI

A commission-free freight marketplace where carriers and shippers transact directly, with AI doing the paperwork.

shipfreight.ai
Ship AI dashboard showing active loads, pending bids, shipment analytics and live tracking
Problem
Freight brokerage takes a cut of every load for work that is mostly document handling and exception chasing. Carriers wanted the load; shippers wanted the capacity; the middle was overhead.
Built
The full platform on Next.js, React, TypeScript, Node.js and Python, plus the money layer end to end: $2M+ moved through Stripe Connect across 4,000+ payouts at 99.9% settlement accuracy, on the transaction model behind carrier payments and shipper invoicing. AI copilots read 12,000+ documents a month at 94% straight-through, triaged exceptions, and let users run logistics work conversationally, always behind a human review gate.
Result
300+ carrier companies onboarded, representing 3,000+ trucks of capacity, tracked against 5 live KPIs — on-time percentage, tender acceptance, dwell, cost per load and exception rate. Event-driven services and WebSockets held 99.9% uptime at sub-200ms push latency underneath it.
TypeScriptNext.jsReactNode.jsPythonPostgreSQLLLM / RAGMCPStripe ConnectDockerAWS
02AI · Verification

Proofbase AI

Verified business metrics pulled straight from the APIs a company already uses, published as a proof page anyone can check.

itsproofbase.com
Proofbase AI homepage showing the platforms it pulls verified metrics from
Problem
Founders, creators and independent operators make claims about their numbers that nobody can verify. Screenshots get taken on faith, and the people doing the vetting have no better option.
Built
10+ one-click OAuth integrations that read metrics directly from source platforms, with zero passwords and zero API keys to manage, then an AI layer that finds patterns across 30+ metric types and surfaces the insight. Output is a public or private proof page with a verified badge on every metric, embeddable anywhere.
Result
Live at itsproofbase.com, refreshing tracked metrics daily for businesses, artists and creators.
TypeScriptNext.jsReactNode.jsPythonPostgreSQLLLMOAuth 2.0Tailwind
03experience

Experience

Nine years between the code and the operation it runs.

May 2025 to Present

Fleet Transportation Dispatcher & Operations

C&S Wholesale Grocers

  • Own the daily delivery plan for a grocery network serving Winn-Dixie, Fresco y Más and Presidente, matching 90+ deliveries a day against available trucks, drivers and dock capacity.
  • Hold 98% on-time performance against contract SLAs across that volume.
  • Redesigned dock scheduling and load release, the process by which work enters the plan, cutting 40 minutes of dwell per load and inter-facility delays 35%, and raising throughput 18% by reallocating routing and labor.
  • Run the recurring operations review that keeps warehouse, receiving and transportation leads aligned on capacity, and on the downstream cost when a date moves.
  • Write a root-cause analysis behind every repeat exception, so the same failure stops coming back.
  • Hold operational records — BOLs, POs, ELD logs — at 97%+ accuracy across integrated systems, so planning, billing and reporting all run off one dataset.
Node.jsPostgreSQLDocker

Jul 2022 to Jul 2026

Operations Technology Lead

Ship AI

  • Built and launched an AI-native, commission-free freight marketplace on Next.js, React, TypeScript, Node.js and Python.
  • Onboarded 300+ carrier companies representing 3,000+ trucks of capacity.
  • Owned the money layer end to end: $2M+ through Stripe Connect across 4,000+ payouts at 99.9% settlement accuracy, on the transaction data model behind carrier payments and shipper invoicing.
  • Shipped AI copilots and agent workflows using LLMs, RAG, MCP integrations, tool calling and structured outputs, letting users run complex logistics work conversationally instead of through forms.
  • Automated document reading and exception triage across 12,000+ documents a month at 94% straight-through, retiring roughly 8 hours a week of manual keying, with a human review gate in front of every action taken and monthly accuracy audits behind it.
  • Designed the capacity model that replaced deal-by-deal guesswork with one integrated forecast, published on a fixed cadence against 5 committed KPIs: on-time rate, tender acceptance, dwell, cost per load and exception rate.
  • Ran the real-time infrastructure underneath it — event-driven services, WebSockets and async processing — at 99.9% uptime and sub-200ms push latency.
  • Owned the design system, CI/CD, Docker and cloud deployment.
Next.jsReactTypeScriptNode.jsPostgreSQLDockerAWS

May 2024 to Jul 2025

Senior Logistics Coordinator

MCI Express Inc

  • Planned and locked capacity for 3 concurrent multimillion-dollar programs — Carnival, Royal Caribbean and Ritz-Carlton — worth $4M+ combined.
  • Committed against fixed vessel and delivery windows, where the plan had to be right the first time because it could not be re-cut.
  • Ran the trade-off calls whenever a date or equipment tranche moved, deciding which program absorbed the slip and pushing the change to accounts the same day.
  • Held external partners to committed dates through a fixed cadence of SLA reviews, escalations and corrective action plans, surfacing risk early enough to re-plan rather than react.
  • Connected transportation platforms and carrier networks through REST and webhook APIs, and integrated EDI exchanges between carrier, broker and shipper systems.
  • Built automated ETL pipelines and standardized reporting across TMS, ELD and GPS telematics in SQL and BI dashboards, retiring a 12-hour weekly manual reporting cycle.
  • Used cost and performance data to find bottlenecks and support pricing and routing decisions.
Node.jsPostgreSQLAWS

Jan 2020 to Mar 2024

Founder & Software Engineer

Dispatch 4 Now LLC

  • Founded and ran an operation that moved $6M+ of freight across roughly 3,000 loads over four years.
  • Owned capacity procurement, rate and contract negotiation, and the margin on every load across a network of 120+ owner-operators and small carriers.
  • Built the internal system of record in TypeScript, React, Next.js, Node.js, Python, PostgreSQL and Redis, covering dispatch, shipment lifecycle, route optimization and driver assignment: the software the business actually ran on.
  • Shipped live dashboards for capacity planning, unit economics, anomaly detection and predictive reporting, and made daily pricing and allocation calls directly off them.
  • Integrated carrier APIs, GPS and telematics providers and mapping services into one operational data platform, tracked in real time over WebSockets.
  • Wrote the SOPs and training library that let 2 dispatchers absorb 4x lane growth without adding headcount.
Next.jsReactTypeScriptNode.jsPostgreSQLGitHub

Jan 2017 to Dec 2019

Bilingual Customer Service Representative

Amazon

  • Supported Amazon’s delivery network at scale against strict service-level targets, working the operational signals behind millions of package events.
  • Traced data synchronization failures across internal platforms back to root cause, validating event consistency through the package lifecycle.
  • Used internal dashboards and monitoring to diagnose delivery anomalies and correlate system events.
  • Recognized as a top performer, working in 2 languages across distributed teams.
AWSLinux
04stack

Stack & capabilities

Press any key on the keyboard. Every keycap is a real one.

Tools

AI Engineering

LLM integrationClaude / Anthropic APIOpenAI APIRAG pipelinesAgent orchestrationMCPTool callingStructured outputsStreaming responsesPrompt cachingEmbeddingsVector search / pgvectorEvals & guardrailsHuman review gates

Languages

TypeScriptPythonSQLJavaScriptHTMLCSS

Backend & Data

Node.jsPostgreSQLRedisREST & webhooksServer ActionstRPCDrizzle / PrismaSupabaseWebSocketsJob queues (BullMQ)ExpressMongoDBFirebaseZodETL pipelinesEDI

Frontend

ReactNext.jsReact Server ComponentsTailwindshadcn/uiRadix UITanStack QueryZustandFramer MotionViteThree.js / SplineGSAPVueSass

Cloud & Infrastructure

AWSDockerKubernetesCI/CDGitHub ActionsTerraformVercelEdge functionsGoogle CloudOpenTelemetrySentryNginXLinux

Finance & Business Systems

Stripe ConnectPayouts & settlementBilling & invoicingReconciliationUnit economicsBI & KPI dashboardsForecasting modelsTMS / WMS / ERP

Tooling & Practice

GitGitHubSystem designCode reviewVitestPlaywrightTurborepoESLintPrettierClaude Codenpm

Hired for

AI & Automation

LLM integration · RAG pipelines · Agent orchestration · MCP · Evals & guardrails · Human review gates

Software Engineering

TypeScript · React & Next.js · Node.js · PostgreSQL · System design

Data & Systems of Record

SQL & data modeling · ETL pipelines · BI dashboards · EDI · TMS / WMS / ERP

Cloud & Infrastructure

AWS & GCP · Docker & Kubernetes · CI/CD · Observability · Edge & serverless

Planning & Operations

Capacity forecasting · Scenario modeling · SOP design · Risk escalation

Payments & Financial Systems

Stripe Connect · Payouts & settlement · Billing · Reconciliation · Transaction modeling

Commercial & Unit Economics

Rate negotiation · Cost-per-unit analytics · Margin & P&L · Pricing strategy

Education & credentials

B.S. Computer Science

2021 to 2024

University of the People. Relational database systems, data analysis, and applied AI/LLM engineering.

NVIDIA Inception

Member. Startup program for AI and accelerated-computing companies.

AWS Activate

Member. Cloud infrastructure credits and technical enablement program.

Languages

Fluent. English, Spanish, and Portuguese.

05contact

Let's talk

Open to roles in software engineering, sales, operations and supply chain. The fastest way to reach me is email.

Email

sebasbecerra70@gmail.com

Elsewhere

LinkedInX

Goes straight to my inbox. Never shared with anyone.