Sean Carroll

Forward Deployed Engineer

Open to full-time roles · Lincoln, NE

I embed with customers, learn their domain fast, and ship production agentic systems inside their stack. Nine years building software, the last six deploying data and AI platforms directly with customers: enterprise marketplace API integrations, GIS/ML underwriting tools built alongside national developers' acquisition teams, and agent workflows, RAG, and evals for 8 startups under HIPAA, SOC 2, and GDPR constraints.

enterprise marketplace API integrations

startups embedded with (avg. 14 weeks)

12M+

parcels evaluated for real estate developers

~65%

LLM spend cut for one client

How I work with customers

The pattern behind every deployment

Discovery with the people who use it

Sit with execs and domain experts, learn how the work is actually done, and write the spec down before writing code.

Ship inside the customer's stack

API integrations into existing product surfaces, tenancy- and PII-aware services, delivery under HIPAA, SOC 2, and GDPR.

Evals and observability from day one

LangGraph agents and RAG over messy corpora, measured against real user feedback instead of demo retrieval.

Hand off something they can run

Runbooks, CI/CD, and a team the customer owns after I leave, including hiring 1–9 engineers per client.

Experience

Customer-facing deployments, most recent first

Co-founder / CTO

Aug 2023 – Mar 2026

BikeWise (SheldonAI) — Exited · Paris, France

·Led customer engineering for 3 enterprise marketplace design partners: scoped each integration with their teams, mapped their catalog data into our schema, and shipped white-glove API integrations into their existing product surfaces.
·Built SheldonAI, an agentic RAG platform over 2.5M+ cycling product listings with CLIP + text hybrid retrieval, reranking, and LangGraph query-time reasoning, tuned against real merchandiser feedback. Took it from zero to EIT Urban Mobility Accelerator recognition.

Fractional CTO / Hands-on AI Engineer

Aug 2023 – Present

Independent · Remote, embedded with client teams

·Embedded with 8 AI-forward startups ($0–$25M ARR) across home health care, real estate, e-commerce, and social; clients include Qensa, ReferraLynx, TrustAgent, and OrBestOffer.
·Shipped LangGraph agent workflows, RAG over messy domain corpora, and eval-driven LLM APIs into each client's existing stack; cut one client's monthly LLM spend by ~65% through model routing, GraphRAG context caching, and prompt trimming.
·Delivered under HIPAA, SOC 2, and GDPR requirements; ran discovery with execs and domain experts and wrote the specs.

CTO

Sep 2020 – Sep 2023

Madelon (REDtech) · New York, NY

·Worked directly with national real estate developers' acquisition teams to learn how they underwrite, then built a GIS/ML platform around that workflow. It evaluated 12M+ parcels across 3 markets and supported 9 deals, with site comparisons in hours instead of weeks.
·Turned messy municipal PDFs, GIS layers, and Excel into reliable PostGIS pipelines; scaled engineering from zero to four with nearshore partners.

Full Stack & Lead DevOps Engineer

2017 – 2020

ADK Group / 3BL Media · Boston & Northampton, MA

·Client-facing full-stack engineering (Drupal, React, Ember); promoted to lead DevOps, standardizing CI/CD, runbooks, and incident response.

Education

Academic background and continuous learning

BS, Materials Science & Engineering

2016

Georgia Institute of Technology

Business Data Analytics Certificate

2018

Wharton School, University of Pennsylvania

AI Program (Master's Certificate)

2026 – present

Stanford University

Projects

Side builds and open-source tools

SwarmHire

Six parallel LangGraph specialists score a resume against a target role and rank what to fix first; ~400 resumes scored against ~2,500 job postings. Next.js, FastAPI/Supabase, Stripe billing.

TAP

The MCP of traceability — a vendor-neutral protocol for signed, verifiable records of AI agent activity.

SUTH

Drives a real browser with LLM-backed personas and reports UX friction, not pass/fail.

Skills

What I use to deploy into customer environments

AI / LLM systems

LangGraphRAG & hybrid retrievalEvals & LLM observability (LangSmith)Multi-agent orchestrationAnthropic ClaudeOpenAIGoogle GeminiCLIPpgvector

Engineering

PythonFastAPITypeScriptNode.jsReactNext.jsSQLPostgreSQLSupabasePostGISRedisDockerAWSGCPGitHub Actions CI/CD

Customer-facing

Technical discoverySolution scoping & written specsAPI integrations into customer stacksDemos to executives & domain expertsHIPAA, SOC 2, GDPR delivery

Let's connect

Looking for a team that deploys AI into the real world.

sean@seacar.ai · +1 (617) 682-0733