The Consulting + Building Balance
Finding the right balance between consulting work and building my own products - how each informs the other.
Roughly 30% consulting, 50% building, 20% learning. Nine months into independence, that's the rhythm that works: consulting for income and connection, building for learning and ownership. In September I said I was not joining another company immediately; nine months is well past immediately.
The Balance
The consulting share is spatial computing strategy for startups, AI architecture for scale-ups, and technical diligence advisory for VCs. Building covers my own products - TrustedChat, the market research system, the bots - plus experiments, prototypes, and open source contributions. The learning slice goes to staying current on AI developments, technical reading and courses, and community engagement.
How Does Consulting Help the Building?
Seeing many companies' problems reveals the common patterns, and client work grounds my building in real market needs instead of my own assumptions. Client relationships also become distribution for what I build. And the boring one: consulting pays the bills while products find traction.
Why Building Helps Consulting
"I built this" carries more weight than "I advise on this." Building keeps my technical skills sharp and keeps surfacing insights I bring back to clients. Having my own revenue also means I can be selective about which clients I take.
What I've Learned
About Consulting
Scope carefully, because scope creep is real. Price on value, not hours. Say no to misfit clients - the energy cost is never worth it - and document everything, since the knowledge compounds.
About Building
Ship small and iterate fast. User feedback beats perfect plans every time I've tested it. Distribution turns out to be as hard as the building itself. And solo has limits; collaboration helps.
About Balance
Calendar boundaries matter, context switching has a real cost, and managing energy beats managing time. I also reassess the split periodically rather than assuming it still works.
Current Status
Products with users right now: TrustedChat (multi-LLM consensus), SwingTrade Pro (trading analysis), the Podcast and News TL;DR bots, and some Instagram automation experiments. On the consulting side I have 3 active engagements, a mix of startup and growth-stage, focused on spatial AI and multi-agent work. Learning time is going to agent frameworks (LangGraph and friends), new model capabilities, and emerging patterns.
Looking Forward
The balance feels sustainable, and it will probably drift toward more building as products gain traction. I've stopped framing it as consulting versus building. The whole thing is one portfolio of activities that compound, and year two of independence is where I find out whether that framing holds.