Skip to content
View akira-in-tech's full-sized avatar

Achievements

Achievement: YOLO

Achievements

Achievement: YOLO

Block or report akira-in-tech

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
akira-in-tech/README.md

👋 Hi, I'm Akira

Software engineer building reliable AI systems and full-stack products.

I care about clear data boundaries, observable workflows, accessible interfaces, and verification as a product feature.

🔎 Evident · 🌐 Live demo · 🗂 Browse my work

🚧 Currently building

Evident, an evidence-driven research workspace that turns complex questions into source-traceable reports through an eight-agent workflow.

🗂 Selected work

Project What it does Explore
✦ Evident Coordinates eight specialist agents across public, academic, MCP, and private-document sources, with durable execution, tenant isolation, citations, and explicit failure states. Repository · Live product · Architecture
🏠 Mortgage Integration Agent Runs parallel mortgage checks through a NestJS and GraphQL service with deterministic provider adapters, PostgreSQL migrations, authentication, rate limits, and integration tests. Repository
🎬 Movie Explorer Delivers movie discovery, reviews, and account workflows through a React and Express product with MongoDB persistence, accessible UI states, tests, and CI. Repository · Live product
🎓 Kanbas Implements a full-stack learning-management experience with React, Redux, Express, MongoDB, session authentication, role-aware authorization, and browser tests. Client · Live client · API

🧭 How I work

  • Make system state visible. Users should understand what is running, what failed, and what they can do next.
  • Treat evidence as part of the interface. Conclusions come first; sources and implementation details remain available on demand.
  • Verify the claim. Tests, evaluation artifacts, health checks, and reproducible environments matter more than impressive-sounding numbers.

🧰 Toolbox

  • 🤖 Intelligent systems: LangGraph, RAG, MCP, evaluation, citations, durable jobs
  • 🧱 Backend: Python, FastAPI, TypeScript, NestJS, GraphQL, PostgreSQL, Redis, MongoDB
  • 🖥️ Product engineering: Vue, React, accessibility, operational states, Playwright
  • 🚀 Delivery: Docker, GitHub Actions, Terraform, AWS, security checks, reproducible test environments

🌱 Open source

I contribute fixes and documentation across financial standards, developer tooling, API clients, authorization, and AI infrastructure:

FINOS FDC3 · Kedro · Stashbase · Voiden · OpenFGA

Pinned Loading

  1. enterprise-agentic-research-platform enterprise-agentic-research-platform Public

    Multi-tenant, evidence-driven research platform: an 8-agent LangGraph workflow combining web, academic, and private-document retrieval with verified citations, built on FastAPI + Vue and deployable…

    Python

  2. mortgage-integration-agent mortgage-integration-agent Public

    Mortgage orchestration reference service with deterministic provider adapters, structured underwriting decisions, NestJS, GraphQL, and PostgreSQL.

    TypeScript

  3. movie-explorer movie-explorer Public

    Deployed movie discovery product with React, Express, MongoDB, accessible UI states, API tests, Playwright coverage, and CI.

    JavaScript

  4. kanbas-react-web-app kanbas-react-web-app Public

    TypeScript React learning-management client with Redux, Vite, accessible product states, Vitest, Playwright, and CI.

    TypeScript

  5. chat-finance chat-finance Public

    Financial research application with Flask, LlamaIndex RAG, FinBERT sentiment analysis, authentication, health checks, tests, and containerized delivery.

    Python

  6. deep-research-agents deep-research-agents Public

    TypeScript and NestJS reference implementation of an eight-agent research workflow with cited report synthesis and tested orchestration boundaries.

    TypeScript