Mock2 blog

1. Client Overview

• Company Name: Triangle
  • LinkedIn: https://www.linkedin.com/company/trianglehealth/about/
• Industry: Software Development
• Stage:
• Location: Pasadena, CA
• Team Size: 10

Brief intro — who they are, what they do, and any relevant context (funding, traction, etc.).

Triangle.AI is developing a platform to revolutionize how patients find and engage with clinical trials. Our system leverages advanced Large Language Models (LLMs) to significantly improve upon existing government websites, offering a more user-friendly, comprehensive, and intelligent approach to clinical trial matching.

• POC
  • https://github.com/ncbi-nlp/TrialGPT

2. The Challenge

What problem were they trying to solve? Why was it important or urgent?

• Context of the business problem

  Triangle Brainstorm

  Core Problem: Clinical trials face massive hurdles in patient recruitment due to poor accessibility, fragmented data, and complicated eligibility criteria.

  Pain Points:

• Existing government databases are hard to navigate and unintuitive for patients and doctors.
• No efficient AI-driven patient-to-trial matching at scale.
• Manual, time-consuming screening processes for trial eligibility.
• Business Need: Deliver an MVP within 3 months that includes key AI-driven matching features while balancing engineering resources.
• Pain points (e.g., lack of engineers, time constraints, performance issues)
  • Spartan need to ensure the accuracy of POC
  • client bring POC Spartan developed to promote to clinical orgs
  • communication: directly with client daily
  • timeline rush
  • dev need to learn clinical business
• Technical challenge:
  • how to know if the system is operating well from a - z
  • lack of data to track if system is good - objective feedbacks from clients → solution: Spartan build air flow system by LLM to run raw data → database for Spartan
  • eligibility matching
• → solution: RFC:
• 958244a9-5d05-4bca-b60d-12b3678dc3ac_RFC-002_Triangle_-_Scaling_The_Search_Algorithm.pdf
• (need to add database of outcomes of trials (articles) → categorize based on effectiveness)
  • time-consuming screening processes for trial eligibility

3. Spartan’s Solution

What did your team do to solve the problem?

• Team composition (e.g., 1 backend + 1 frontend engineer):
  • 3 devs (2 AI, 1 full stack)
  • client: 5 inhouse engineer (interns)
• Approach (tech stack, processes, integrations)
• Key deliverables
• Collaboration style (e.g., async daily standups, tight founder-engineer loop):
  • Slack channel
  • Notion: manage tasks
  • Daily Meeting to report to client
  • Github
• POC that changed perspectives of client (e.g. eligibility matching)
• Built the MVP based on the TrialGPT framework, integrating zero-shot patient-to-trial matching.
• Developed an AI-powered search interface + chat interface for trial discovery.
• Built operational dashboards for managing patients, trials, and AI matching processes.
• Created an automated pag e generation pipeline combining scraped trial data, PubMed insights, and LLM-generated summaries.
• Enabled multilingual support to serve global users.
• Designed privacy-preserving features like anonymous patient profiles and trial ownership claims.

4. The Outcome

What was the result of working with Spartan? Be specific and metric-driven.

• MVP shipped in X weeks
• % faster time to market
• Increased user retention/engagement
• Helped them raise funding
• Tech debt reduction / scalability improvements
• client demo to billionaires (cancercommon - https://cancercommons.org/)
• Triangle AI shipped a functional MVP that:

  → Enables users to search through 400,000+ active clinical trials globally.

  → Uses AI to deliver 87.3% accuracy in patient eligibility matching, near expert-level performance.

  → Reduces trial screening time by 42.6%, significantly streamlining patient recruitment.

  → Provides multilingual access and a user-friendly interface far superior to existing government platforms.

• Positioned Triangle AI to move into fundraising and partnership discussions with medical institutions and trial sponsors.

5. Client Testimonial (if available)

Justin Balthrop - Co-founder & CTO of Triangle, Co-founder of Bird

https://www.linkedin.com/in/justinbalthrop/

Spartan’s founder, Chan, is one of the best engineers I’ve ever worked with—I used to joke that I wanted to figure out how to clone him. With Spartan, it feels like he’s done exactly that. Their hiring bar is incredibly high, and the developers I’ve worked with are all at the same high level, consistently delivering outstanding work with impressive speed.”

6. Tech Stack

List of core technologies used.

• Backend: Python, Node.js
• Frontend: React, Next.js
• AI/LLM: OpenAI API, LangChain, Pinecone/Weaviate for vector search
• Infrastructure: AWS (Lambda, S3, DynamoDB), Docker, Terraform
• Data Sources: WHO Trial Database, PubMed, ClinicalTrials.gov

7. Why Spartan

Short section explaining why the client chose Spartan.

• Fast onboarding
• Startup-native culture
• Engineering craftsmanship
• Cost-efficient vs SF hiring
• Deep AI expertise: Experience deploying production-grade LLM applications.
• Startup-native speed: Rapid prototyping with MVP shipped in under 3 months.
• Product + Engineering craftsmanship: Balanced between delivering user value and technical scalability.
• Cost-effective team augmentation: Access to elite engineers without SF/NY costs.

🔁 Optional Sections

• Before/After product screenshots
• Timeline graphic
• Links to live product / app