Technical founder. Co-founder and CTO at Akisto.
I build AI systems for work where mistakes are expensive.
For a decade I have taken products from first idea to production, across applied AI, developer infrastructure, identity, and real-time systems. In every company I have owned the technical foundation, from the first architecture decision to production.
Akisto, co-founder and CTO. An AI supplier coordinator for manufacturing procurement teams. Between a purchase order and its delivery, buyers chase suppliers through email and calls that mix several orders, answer half the question, and move dates without saying so. Akisto automates that work: traceable facts from every conversation, a consistent state for every order, and follow-up that happens on time. It is deployed with a manufacturer and has managed more than 2,000 purchase orders.
I lead product and engineering end to end: system architecture, AI agents that work with suppliers over email and phone, workflow automation, integrations, data pipelines, and a risk layer that flags orders likely to disrupt production. TypeScript, React, PostgreSQL, Temporal, AWS.
| Years | Company | Role | What it was |
|---|---|---|---|
| 2026 | Akisto | Co-founder and CTO | AI supplier coordination for manufacturing procurement. Deployed with a manufacturer; 2,000+ purchase orders managed. |
| 2025–26 | Scrute | Co-founder and CTO | AI document intelligence for commercial real estate. Title commitments reconciled against ALTA surveys, plan sets checked against building-code rules, every finding tied to its source. Customer discovery moved the focus from title review to permit compliance. Demos |
| 2025 | Commentrix | Founder | Real-time AI sports commentary. A working MVP in two weeks; market research showed too few buyers to build a company on, so I chose not to pursue it. Source |
| 2025 | Aiditor | Co-founder | A conversational editor for image, audio, and video. I built the complete product from scratch, through to launch readiness. Source |
| 2023–24 | Authnest | Founder | Built solo through several iterations, from social login to enterprise authentication to a dedicated email service for logins, one-time codes, and notifications. Source |
| Since 2016 | Mamidi Ventures | Founder | My independent studio: consumer software, games, SaaS, AI, and developer tools, each taken from zero to one. Aftlog, Everby, SlowAndSteady, Rank Race, SV Hunt, Twish, OnlyOnePremium, Links Browser, Favs.bio, Vowalls. |
Aftlog: a flight recorder for AI coding agents. A local Rust daemon records every command and file change, separates the agent's work from what was already there, and produces a readable report with a conservative, backup-first revert plan. Rust, Git, filesystem watchers.
Scrute: document review that cites its sources. Title and survey parsing run in parallel and are reconciled together; plan details from a vision model are checked by explicit TypeScript rules; findings carry the rule, observed and required values, and page references. TypeScript, NestJS, PostgreSQL, Gemini.
Aiditor: conversation as the control surface for media. The model plans the edit as Python, a dedicated container executes it, and failures feed back into the next attempt. Python, FastAPI, Gemini, Docker, FFmpeg.
Everby: a native macOS accountability companion. Layered memory with distinct lifetimes, check-ins timed to the person's actual day, and personal state that stays on the machine. Swift, AppKit, Keychain, tool calling.
- Real-time systems. SlowAndSteady.io, a zero-player multiplayer race that drew around 30,000 visits in one night, and SlowAndSteady 2, a shared economy where giving creates power. WebSockets, Redis, Unity.
- Multimodal pipelines. Commentrix samples footage, writes timestamped commentary, and aligns narration and subtitles to the source timeline.
- Identity and email. Authnest, SES Email Ingestion, and an OAuth Flow Tester for proving authorization-code and OIDC flows in isolation.
- Browser automation at scale. Twish, a visual index of one million websites captured by parallel Puppeteer workers on EC2, and the Ad Network Scanner.
- Explainable decisions. The Logistics Trust Console, carrier due diligence that refuses to hide its reasons.
More: Open Order, Voice Chat Widget, Prompt Page Builder, WhoCan, AI Wallpaper Batcher, OnlyOnePremium.
Start where the work actually happens. Before any architecture, learn how the job is done today: the spreadsheets, the email threads, the workarounds.
Let models interpret. Let rules decide. Models read messy input well. Commitments, state, and money belong to deterministic code, with a person able to step in.
Show the evidence, and plan the undo. A finding, score, or automated action is incomplete without a source someone can inspect and a way back if it is wrong.
Earn every piece of complexity. Use the smallest architecture that tells the truth.
Working on a difficult problem in a domain you know deeply? I would like to hear about it.