The infrastructure, end to end.
AgentForge is a production multi-agent operating system that runs four of my business units unattended — sourcing, diligence, modeling, outreach, and reporting. This is the full stack: the agents, frameworks, router, tools, memory, evals, and automations. Built and maintained solo.
One request, seven layers.
Every interaction flows the same path — capture, orchestrate, route, execute, remember, score, report — each layer independently observable and fault-tolerant.
Intake & routing
Universal capture from Telegram, Fireflies, and HTTP → semantic dedup (Voyage embeddings, 0.93 cosine) → structured classification → auto-route at ≥0.85 confidence, else a human-review queue.
Orchestration
A CEO-supervisor agent decomposes the request into a MECE task DAG, then dispatches to the right team — wrapped in a 9-layer middleware stack (caching, retries, circuit breakers, cost tracking, call limits).
Agent teams
Six specialized teams (20 agents) execute — consulting & PE, coding, strategy, opportunity discovery, LBO modeling, introducer enrichment — with parallelized independent sub-tasks.
Model router
A 9-tier cost-optimized router picks the cheapest model that meets the bar (Groq → Gemini → MiniMax → Claude), with automatic circuit-breaker fallback per tier.
Tools
44+ tool definitions across 56 modules — research, enrichment, financial modeling, outreach, dev, media — invoked under per-run tool-call and cost limits.
Memory & knowledge
Dual persistence — Mem0 semantic memory + a Graphiti/FalkorDB temporal knowledge graph — plus hybrid RAG (Qdrant vectors + BM25 + cross-encoder rerank) over ingested documents.
Evals, observability & report
Per-section QA gates, an independent test agent, and full Langfuse tracing (cost, latency, tokens) on every run — then results report back to the operator.
Modern agentic infrastructure.
The frameworks and services the system is actually built on — not a wish-list.
Frameworks & runtime
Memory, knowledge & data
Models & observability
Twenty agents. Six teams. One operator.
Consulting & PE
Does: URL extraction → research → market intel → company profile → deal score → PE value-creation plan → personalized approach letter. Benefit: per-section quality gates (8+/10 to pass) and a parallelized MECE DAG — a full diligence pipeline for $2–5.
Coding
Does: ships features end-to-end; the test agent reads the PRD (not the code) to prevent coverage theater; the Review/Heal loop runs tsc + biome and self-corrects up to 5 iterations. Benefit: coherent, validated code at MiniMax cost.
Opportunity discovery
Does: scans 40+ verticals for weak signals, finds cross-domain patterns, scores on 5 dimensions (readiness, competition, buildability, scalability, defensibility). Benefit: spots trends before mainstream awareness — ~$1.50 a scan.
LBO financial modeling
Does: deal intake → Docling/Opus extraction → quality-of-earnings → SBA prescreen gate → formula-driven 17-tab 3-statement Excel → 3-scenario sensitivity. Benefit: IC-grade models in under five minutes; 50+ QA-validated.
Introducer enrichment
Does: discover → enrich (Firecrawl/Apollo) → paid-tier deep → web-search → validate → recover → sync. Benefit: verified 56% email coverage — 4,100+ deliverable contacts from a 7,000+-firm cohort.
Intake routing
Does: turns voice, photos, and notes into deduplicated, classified, confidence-scored work items. Benefit: nothing is dropped, nothing is processed twice, and only the uncertain reaches a human.
The cheapest model that clears the bar.
A 9-tier router across 8 providers, each tier with a circuit-breaker fallback. Caching cuts repeat-prompt cost ~90% (~$270/mo verified).
44+ tools the agents actually call.
8 tools
- Perplexity — grounded web research + citations
- CellCog — multi-source synthesis
- Exa — semantic company search
- Tavily / SerpAPI — fast web + SERP
12 tools
- Apollo — org + people, paid-tier reveal
- LeadMagic / Datagma — email find + validate
- Firecrawl — scrape + typed extraction
- Google Places · PDL — discovery + scoring
5 tools
- LBO generator — 17-tab formula-driven Excel
- Financial extraction — Opus over CIM/K-1
- Tax engine — W-2 + S-Corp, 25 strategies
- EBITDA / revenue — sizing for screening
3 tools
- Mem0 — long-term semantic facts
- Knowledge graph — temporal entities
- MegaRAG — hybrid retrieval + citations
4 tools
- Microsoft Graph — email from own domain
- Smartlead — campaign sequencing
- Resend — transactional mail
- Twilio — SMS / voice (optional)
11 tools
- GitHub — clone, commit, push, deploy
- Vercel — auto-deploy on main
- File + bash — via Claude Agent SDK
- Supabase · Qdrant · Langfuse — data + traces
Where it does real work.
Verified, in production — not demos.
Industry research → lead discovery
A daily AI thesis → 20 deep sub-niches → a Google Places / Exa / Apollo / Firecrawl discovery waterfall. Verified: 631 leads from one thesis in 16 min 33 s.
Introducer enrichment
Discover → enrich → paid-deep → web → verify → recover → sync, with provider-error isolation. Verified: 4,100+ deliverable emails (56% coverage of a 7,000+-firm cohort).
LBO modeling
Documents in → quality-of-earnings → SBA prescreen → 3-statement model → sensitivity → cross-check. Verified: 50+ IC-grade models, formula-driven.
Coding team
Feature request → PRD → architecture → code → independent tests → tsc/biome review → self-heal. Benefit: validated features at draft-model cost.
9-layer middleware on every agent.
Summarization · tool retry · tool-call limit — survives long runs and transient failures; caps runaway chains at 30 calls.
Model-call limit · model retry · prompt caching — caps per-run cost, auto-recovers 429/500/503, and caches for ~90% savings.
Filesystem · sub-agent spawn · cost tracking — safe I/O, asyncio parallelism, and per-call cost logged to Langfuse.
Nothing ships unscored.
Per-section quality gates — each artifact scored 1–10, 8+ to pass, max 2 rewrites.
Independent test agent — reads the PRD, not the code, so it can't game coverage.
Semantic dedup — Voyage embeddings, 0.93 cosine, before anything is stored.
Langfuse — cost, latency, tokens, and full traces on every run.
19 Claude Code skills — reusable, versioned.
One-off agent workflows turned into products the system can invoke on demand.
Four business units, run by the same system.
Paradosi Partners
- Sourcing, enrichment, scoring, and outreach for a lower-middle-market industrial search — 18,000+ leads and 7,000+ referral relationships under management.
ClearForge.AI
- Strategy factory + SightForge sales intelligence + LBO modeling delivered as client engagements.
Investment scanner
- Auction and foreclosure scanners with per-property investor diligence and automated deep research.