Senior Technical Developer

Senior Technical Developer (Contract)

Company: MiraModo Inc.
Position: Senior Technical Developer (Contract)
Commitment: 25–40 hours/week
Location: Remote (Global)
Timezone: Must overlap with U.S. Central/Eastern daytime hours
Reports To: Founder

MiraModo Inc. is looking for a Senior Technical Developer to become the founder's primary implementation partner.

You will work directly in a production codebase serving enterprise clients. The founder handles product direction, AI architecture, and client relationships — you own the building.

This role combines full-stack development, ML/AI infrastructure, and specialized construction project controls.

There are no management layers or unnecessary ticket bureaucracy. You'll work directly with the founder to design, build, test, and deploy production systems.

Months 1–2: Guided

• Founder provides task specifications.
• You design, implement, and submit pull requests.
• Founder reviews and provides feedback.

Months 3–4: Semi-Autonomous

• Take broader feature briefs.
• Propose technical approaches and architecture.
• Handle production issues with less oversight.

Month 5+: Trusted Technical Partner

• Independently scope and deliver features.
• Propose technical improvements.
• Contribute to technical direction and architecture decisions.

• Detailed architecture and system documentation will be provided to selected candidates under NDA.

TypeScript

• 3+ years of production experience
• Strict mode
• Generics
• Discriminated unions
• Type inference

Python

• Production proficiency
• ML pipelines
• Training
• Evaluation
• Data preparation

React

• Production experience
• Hooks
• Component architecture
• Latest React patterns
• Must have shipped and maintained real applications

Next.js

• App Router
• Server vs. client components
• Route handlers
• Middleware

PostgreSQL

• Strong SQL skills
• Complex queries
• CTEs
• Window functions
• Schema design
• Migrations
• Raw SQL — no ORM dependency

Type-Safe APIs

• End-to-end type safety between client and server
• Experience with tools such as tRPC

Git

• Clean commits
• Feature branches
• Pull-request-based development
• Code review discipline

Tailwind CSS

• Utility-first CSS
• Responsive design
• Data-dense layouts

Component Libraries

• Headless or unstyled component primitives
• Experience with libraries such as Radix

Data Visualization

• Charting libraries
• Dashboards
• S-curves
• Trend charts

Report Generation

• PDF
• Excel
• Word
• PowerPoint exports

PyTorch & Hugging Face

• Model fine-tuning
• PEFT/LoRA
• Transformers

Voice / Speech ML

• Speech-to-text (STT)
• Text-to-speech (TTS)
• End-to-end voice pipeline development

GPU Infrastructure

• Self-hosted GPU servers
• CUDA
• Resource monitoring

Inference Serving

• High-throughput model serving
• Quantization
• Performance optimization

Evaluation

• Standard evaluation frameworks
• Custom, domain-specific evaluation harnesses

Docker

• Containerization
• Multi-stage builds
• Docker Compose

CI/CD

• Automated build
• Testing
• Security scanning
• Deployment workflows

Self-Hosted Deployment — Bonus

• VPS-based hosting
• Container orchestration
• Monitoring

Testing

• Unit testing
• Integration testing
• Testing frameworks

• Strong TypeScript combined with strong ML skills is a rare combination.
• A candidate with excellent TypeScript/React/SQL fundamentals and a demonstrated ML learning trajectory — including personal projects, coursework, fine-tuning experiments, or similar work — can be an excellent fit.
• Aptitude and drive matter more than arriving with deep ML expertise.

• Our professional domain is capital construction project controls — measuring, forecasting, and managing cost and schedule performance on large construction programs.
• You do not need prior construction experience. However, you must be willing and able to learn the domain.

• Earned Value Management
  Industry-standard methodology for measuring project performance by comparing planned work, completed work, and actual cost.
• Schedule Analysis
  Critical path analysis, schedule forecasting, delay identification, and forensic scheduling.
• Field Progress Measurement
  Objective methods for quantifying physical completion using predefined measurement criteria.
• Cost Management
  Budgeting, cost tracking, forecasting, variance analysis, and S-curve trend visualization.
• Change Management
  Scope change tracking, information requests, deficiency management, and their impact on cost and schedule.
• Work Packaging
  Organizing construction activities into executable packages across engineering, procurement, and construction.
• Each of these concepts maps directly to platform features.
• Domain knowledge is taught during onboarding through documentation, guided tasks, and founder mentorship.

• Hours:
  25–40 hours/week, with close to full-time commitment expected.
• Location:
  Global candidates welcome.
• Timezone:
  Must be available for meaningful overlap with U.S. Central/Eastern daytime hours, approximately 9:00 AM–5:00 PM CT.
• Communication:
  Async daily communication through Slack/Discord, with video calls as needed for architecture discussions.
• Code Workflow:
  Feature Branch → Pull Request → Founder Review → Merge → Automated Deployment
• Tooling:
  AI-assisted IDE, GitHub, Docker, PostgreSQL client, and modern development tools.

• First 30 Days
• Environment fully running
• 5–8 pull requests merged
• Able to trace data from database through the application to the UI
• Basic understanding of the construction project controls domain
• First 60 Days
• Independently implementing features
• Confidently writing SQL migrations
• Handling production bugs
• Completing initial ML fine-tuning experiments
• First 90 Days
• Decomposing feature briefs into actionable tasks
• Proposing architecture
• Comfortable across the full stack
• Contributing to voice pipeline and ML infrastructure

• 3+ years of production TypeScript/JavaScript experience
• Strong understanding of strict TypeScript, generics, and type inference
• Python proficiency for ML pipeline work, including training, evaluation, and data preparation
• Production React experience
• Experience shipping and maintaining real applications
• Strong understanding of Next.js App Router
• Strong SQL skills using raw PostgreSQL
• Experience with or strong interest in LLM fine-tuning and ML infrastructure
• Clean Git workflow and pull-request-based development discipline
• Clear written English for code reviews, documentation, and occasional client communication
• Ability to learn independently within a specialized domain and rapidly evolving ML engineering environment

• Candidates with any of the following will stand out:
• Hands-on PyTorch/Hugging Face model fine-tuning
• GPU infrastructure management
• CUDA experience
• Voice/speech ML, including STT, TTS, or voice pipelines
• Inference optimization and quantization
• High-throughput model serving
• Data visualization and chart-heavy dashboard development
• Docker and CI/CD
• Container-based deployment
• Construction, engineering, or industrial domain experience

• Direct Impact
  Your code will serve live clients working on active capital construction programs.
• Production ML
  Fine-tune LLMs, build voice pipelines, and manage inference infrastructure. This is real production ML — not demos.
• Domain Depth
  Learn a specialized, high-value field that most software developers never encounter.
• Growth Path
  Grow from guided implementation to becoming a trusted technical partner.
• Modern Technology Stack
  Work with current technologies applied to real-world problems rather than basic CRUD applications.
• Founder-Direct Collaboration
  Work directly with the founder. No unnecessary management layers and no ticket theater.

• Please submit the following:

• Provide 2–3 paragraphs explaining:
• Who you are
• Why this role interests you
• Your most relevant experience

• Please include relevant:
• GitHub
• Portfolio
• Production applications
• Technical projects
• ML/AI projects
• Sample Demo

• Please include:
• Weekly availability
• Timezone
• Earliest possible start date
• Rate expectations

• Please answer in 4–8 sentences.
• You have a PostgreSQL table:
• work_items (id, category_id, discipline, budgeted_hours)
• and a table:
• progress_entries (id, work_item_id, milestone_index, completed_at, credit_percent)
• Write a SQL query that calculates total earned hours per discipline for a given category_id.
• Explain your approach.

• Please answer in 4–8 sentences.
• You are fine-tuning an LLM for a specialized technical domain where the base model frequently produces incorrect terminology.
• Describe your approach, including:
• Training data preparation
• Fine-tuning strategy
• How you would evaluate whether the model has measurably improved

• Selected candidates will receive detailed technical documentation under NDA during onboarding.