pro
Yonatan Tesfaye
AI Automation Engineer | n8n, Agents & Integrations
New to Contra
Yonatan is ready for their next project!
Nova4AI Command Center
Most AI agent products give you one assistant and a text box. Nova4AI gives you a fleet you can watch work.
Ten specialists run as a team a Researcher, a Coder, a Designer, a QA reviewer, an Archivist, a Prospector, an Optimizer with Navi orchestrating: routing tasks, running standups, pulling teammates into threads with an @mention. Every agent takes voice input.
The core of the build is the Live Feed. Agent-to-agent handoffs stream in real time, each carrying a task ID, trace ID, source system, and timestamp. When Kenzo passes work to Relay, you see the message and can follow the trace. Multi-agent systems usually fail silently and invisibly; this one is auditable by default.
Around the fleet sits the actual business surface: CRM and lead pipeline, proposals, performance analytics, projects, a shared memory layer, docs, and a skills registry. Agents write real artifacts to real endpoints and integrate outward through MCP and n8n.
At the time of this screenshot: 10,983 activities in a day, 5,017 feed messages, 110 tasks open and 42 closed.
Built so you can hand work to a team and still know exactly what happened.
0
22
Fundory — Grant Intelligence for Nonprofits
Grant funding runs on a brutal arithmetic: most applications fail, and the ones that fail still cost weeks of a small team's time. Nonprofits with two or three staff are spending that time writing into opportunities they were never positioned to win.
Fundory attacks the problem upstream. Instead of helping organizations write faster, it decides what's worth writing at all. Every discovered grant is screened against the organization's actual profile and returned with a fit score, so a 65% never consumes the effort a 91% deserves. The product states its threshold plainly: apply above 85%.
The pipeline runs Discovered → Screened → Waiting for Approval → Drafted → Closed, with AI moving work forward and humans holding the approval gate. Drafted proposals carry two separate scores — model confidence and draft quality — because those measure different risks and collapsing them would hide the one that matters.
I built the screening and scoring layer, the document and narrative system that grounds drafts in the organization's real history, and the pipeline that keeps 28 live opportunities legible to a team without a grants department.
Honest signal over generated volume.
2
38
Bench — AI Agents for Small Teams
Solo founders don't lack AI tools. They lack the people who would use them. Most products hand you a blank text box and quietly make you the operator so the inbox still doesn't get triaged and the content calendar still doesn't get built.
Bench inverts that. Instead of a chat interface, it's a hiring interface. You staff a bench of role-trained specialists — an Executive Assistant, a Lead Generator, a Social Media Manager, a Marketing Manager — each scoped to real responsibilities with its own tool access and definition of done. They work in the background and report status: working, done, needs review.
I built the agent orchestration behind it, the role definitions that keep each specialist in its lane, and the review layer that surfaces anything consequential before it ships. Drafts land in the founder's voice. Nothing gets sent without a human seeing it first.
The design constraint throughout was trust. Multi-agent systems fail when they act confidently in the wrong direction, so the product is architected around visible work and cheap correction rather than silent autonomy.
Built for the smallest teams to deploy what used to require a payroll.
1
35
An n8n pipeline that monitors prospect social accounts, decides what's worth responding to, drafts on-brand comments with Claude, and posts them through a real browser session with human review built in.
0
36