Services
Software, AI, and data work — built end to end.
I design, build, and ship production systems: Django web platforms, Flet mobile apps, AI agents, MCP integrations, fine-tuned models, and data tools. Each engagement is scoped to your problem and budget.
Web Platform (Django)
Full-stack web applications — from idea to deployed product.
- Django backend + REST API
- HTMX or templated frontend, Tailwind styling
- Auth, payments (Paystack), and role-based access
- PostgreSQL / MySQL / SQLite (WAL)
- Deployment on Linux, Render, or your cloud
Cross-Platform App (Flet)
One Python codebase shipping to Android, iOS, Windows, macOS, Linux, and web.
- Flet app — true cross-platform from one codebase
- Offline-first SQLite (WAL) storage
- Sandboxed local runtime (pandas, matplotlib)
- AI gateway integration (multi-model with failover)
- AdMob monetization + credits system
- Google Play submission with APK splits
AI Agent Pipeline (Google ADK)
Autonomous agent systems that research, analyze, and act.
- Sequential / loop agent orchestration on Google ADK
- Writer–critic refinement loops
- Vision, audio, and text multimodal analysis
- Web research (DuckDuckGo, search APIs)
- MCP client integration to your data sources
MCP Server or Integration
Make your product's data available to AI agents and tools.
- Custom MCP server (Python, published to PyPI if needed)
- GitHub Tree sync, diskcache, fuzzy search
- MCP client wiring into existing pipelines
- Client configs for Cursor, Claude, VSCode, Zed
Model Fine-Tuning
Fine-tune open models for your domain or language.
- LoRA / PEFT fine-tuning (Phi-3, Llama, Gemma, etc.)
- Dataset prep, formatting, and cleaning
- Checkpoint-resume strategy for long training runs
- GGUF export + publish to Hugging Face
- Inference scripts (llama.cpp, Ollama)
Vibe Code Audit
AI-generated codebases cleaned up, hardened, and made production-ready.
- Full audit of AI / LLM-generated codebases
- Bug hunting, dead code removal, refactoring
- Security review and dependency fixes
- Tests, types, and documentation gaps
- Clear fix list + actionable report
Data Tools & Dashboards
Collect, analyze, and visualize data — privacy-first if you need it.
- Data ingestion (CSV, Excel, SQL, JSON, APIs)
- Pandas / NumPy analysis pipelines
- Interactive charts and reports
- Survey/form generation + live responses
- Export to PDF / PowerPoint
Frequently Asked Questions
Email me with a short description of your project, your goal, your timeline, and a rough budget. From there I'll send a scope and proposal.
Yes. After launch we can arrange a monthly retainer for updates, monitoring, and support to keep things secure and fast.
A data tool or MCP server is often 1–2 weeks. A Django platform or Flet app usually runs 3–8 weeks depending on scope. AI agent pipelines and fine-tuning vary most — I'll give a precise estimate up front.
Yes. I'm comfortable onboarding into existing Django, Flet, or Python projects — adding features, refactoring, or integrating AI without a full rewrite.
If you want it. I can package a tool for PyPI, submit a Flet app to Google Play (with APK architecture splits), or publish models and datasets to Hugging Face under your account.
Have a project in mind?