Founding AI Engineer - Agentic Systems & LLM Ops

Brandon Lin

Houston, TX • **************@*******.*** • 713-***-****

SUMMARY

Founding AI Engineer with 10+ years of experience building production AI systems, agentic workflows, and distributed architectures. Proven track record of architecting and deploying LLM-based solutions, multi-step reasoning pipelines, and evaluation frameworks that improve intelligence and reliability. High-agency leader who ships fast in ambiguous environments, driving measurable gains in system accuracy and user outcomes. Ready to own core systems, collaborate directly with founders, and advance the frontier of financial AI.

EDUCATION

University of Texas at Dallas Richardson, TX

Bachelor’s Degree in Computer Science August 2011 - May 2015

PROFESSIONAL EXPERIENCE

ScienceSoft

Senior AI Engineer November 2023 – Present

Architected and deployed a multi-agent orchestration system using LangGraph and FastAPI, reducing task completion time by 40% across 5 concurrent agents.

Designed and implemented a knowledge retrieval pipeline with Pinecone and LLM embeddings, improving answer accuracy by 25% for complex queries.

Led development of an LLM-based code generation tool that writes and executes Python scripts autonomously, boosting research iteration speed by 60%.

Built evaluation loops with Pytest and custom metrics, enabling continuous model improvement and reducing hallucination rates from 12% to 4%.

Collaborated with product managers and founders to ship 3 major releases in 8 months, each enhancing reasoning capabilities and user satisfaction.

Mentored 4 junior engineers on agent architecture, code reviews, and best practices, increasing team delivery velocity by 30%.

Owned end-to-end deployment of AI services on AWS using ECS and Lambda, ensuring 99.9% uptime and scalability under variable load.

Integrated observability tools (Prometheus, Grafana) to monitor agent performance and system health, enabling rapid incident response.

Developed a distributed event-driven pipeline for real-time financial data ingestion, processing 50,000 events per second with sub-second latency.

Designed and implemented a production-grade API layer using FastAPI to expose AI agent capabilities to internal and external clients.

Led the transition from monolithic to microservices architecture, improving system modularity and reducing deployment failures by 50%.

Created automated testing frameworks covering unit, integration, and end-to-end tests, achieving 85% code coverage across AI services.

Optimized Docker and Kubernetes deployments for AI workloads, reducing infrastructure costs by 35% while maintaining performance.

Conducted technical interviews for AI engineering roles, evaluating candidates on Python, system design, and LLM expertise.

Authored comprehensive documentation covering system architecture, API specifications, and operational runbooks for AI pipelines.

Drove innovation by researching and prototyping new AI techniques, presenting findings to stakeholders and influencing product roadmap.

Innos

Lead AI Engineer July 2022 – October 2023

Architected and built an AI agent platform using LangChain and LlamaIndex, enabling autonomous data analysis and report generation for finance clients.

Designed multi-step reasoning workflows that iterated on results, achieving a 30% improvement in prediction accuracy over baseline models.

Led a team of 5 engineers in developing an LLM-powered system for answering complex financial questions, processing 1,000+ queries daily.

Implemented evaluation pipelines with custom metrics to measure reasoning quality and reliability, reducing system errors by 20%.

Deployed scalable AI infrastructure on AWS SageMaker and ECS, supporting rapid experimentation and model iteration cycles.

Collaborated with founders and product managers to define product requirements and technical vision for AI capabilities.

Developed a knowledge retrieval system using ChromaDB and embedding models, improving factuality and reducing incorrect outputs by 15%.

Created automated code generation and execution modules, enabling the system to write, run, and validate data-science code in real time.

Built observability dashboards with Grafana and custom logging, providing real-time visibility into agent behavior and performance.

Mentored junior engineers on agent architecture, Python best practices, and testing strategies, fostering a high-performance team.

Owned the full software delivery lifecycle from architecture to deployment, shipping 2 major product releases within tight deadlines.

Designed and implemented a caching layer using Redis, reducing response latency by 40% for frequently accessed queries.

Conducted code reviews and enforced coding standards, improving code quality and reducing bugs by 25%.

Software People Inc

Senior Backend Engineer April 2020 – June 2022

Designed and built high-throughput RESTful APIs using FastAPI and PostgreSQL, supporting 10,000 requests per minute with 99.9% uptime.

Developed event-driven data pipelines processing 500,000 messages daily using AWS Lambda, SQS, and S3 for real-time analytics.

Led migration of monolithic backend to microservices architecture, reducing deployment time by 60% and improving system reliability.

Implemented automated CI/CD pipelines with GitHub Actions and Terraform, enabling zero-downtime deployments and infrastructure as code.

Collaborated with QA engineers to build comprehensive test suites using Pytest and Selenium, achieving 90% code coverage.

Optimized SQL queries and database indexing, reducing query execution time by 40% for heavily accessed endpoints.

Participated in on-call rotation, resolving production incidents within SLA and improving mean time to recovery by 30%.

Documented architecture decisions and operational procedures, enabling smoother knowledge transfer and team scaling.

IT Vision Group LLC

Backend Engineer September 2017 – March 2020

Developed RESTful APIs using Python Flask and PostgreSQL for a financial analytics platform, handling 5,000 concurrent users.

Designed and implemented data models and ETL pipelines to aggregate market data from multiple sources, improving data freshness by 50%.

Built automated tests with Pytest and integrated with CI/CD, reducing regression bugs by 30% and accelerating release cycles.

Collaborated with cross-functional teams to define API specifications and ensure alignment with business requirements.

Participated in agile ceremonies and code reviews, contributing to a culture of continuous improvement and code quality.

KoderLabs

Junior Backend Engineer June 2015 – August 2017

Designed and built RESTful APIs using Python Flask and PostgreSQL for a SaaS platform, supporting 2,000 daily active users.

Developed database schemas and optimized queries, reducing page load times by 35% and improving user experience.

Implemented unit and integration tests with Pytest, achieving 80% code coverage and reducing production bugs by 20%.

Participated in agile development processes, collaborating with frontend engineers to integrate APIs seamlessly.

Documented API endpoints and system architecture, aiding new team members in onboarding and project continuity.

SKILLS

Languages:Python, SQL, JavaScript, TypeScript

Frameworks & Libraries:LangChain, LangGraph, LlamaIndex, FastAPI, Flask, PyTorch, TensorFlow, Hugging Face Transformers, React

Cloud & DevOps:AWS (Lambda, SageMaker, ECS, EC2), GCP, Docker, Kubernetes, Terraform, CI/CD (GitHub Actions, Jenkins)

Databases:PostgreSQL, MongoDB, Redis, Pinecone, ChromaDB, Weaviate

APIs & Architecture:RESTful APIs, GraphQL, Event-Driven Architecture, Microservices, Agent Architectures, Multi-Step Workflows

Testing & Quality:Pytest, UnitTest, Selenium, Evaluation Pipelines, Observability (Prometheus, Grafana), Logging

Tools & Methodologies:Git, JIRA, Agile, Scrum, Code Reviews, Mentoring, Documentation, Production Support

PROJECTS

Financial AI Agent Platform (ScienceSoft)

Architected and built an AI agent platform using LangGraph, LangChain, and FastAPI that autonomously answers complex financial questions by writing and executing Python code, and iterating on results. Designed multi-step reasoning workflows and deployed on AWS with Docker and Kubernetes. Achieved 40% faster insight generation and 25% higher accuracy compared to manual analysis.

Autonomous Financial Research System (Innos)

Led development of an LLM-powered system for financial research using LangChain, LlamaIndex, and AWS. The system retrieves knowledge, writes data-science code, executes it, and iterates to generate investment insights. Built evaluation loops and observability dashboards to ensure reliability. Reduced query response time by 30% and improved prediction accuracy by 20%.

Contact this candidate