Ai-Con.

The problem

Traditional consulting services are prohibitively expensive and structurally biased toward large clients — placing serious strategy work out of reach for most small and medium-sized energy enterprises. SMEs in oil & gas, utilities, and renewables face the same forces shaping the majors (energy transition, regulation, capital allocation, technology choice) without the budget or in-house strategy team to navigate them.

The compounding pain: SME clients often struggle to articulate their business context clearly enough for any consulting input to be useful, struggle to define scope when they don't know the consulting frameworks, and distrust black-box AI with strategic recommendations — especially when the cost of a wrong call is operationally enormous in capital-intensive industries.

The solution

Ai-Con replicates the structure of a real consulting team in software — multi-agent collaboration with defined roles, mirroring how actual MDs, EMs, Senior Associates, and Analysts work together to produce strategy deliverables. The output is end-to-end: market analysis, competitive benchmarking, strategic recommendations, and energy-transition advisory, all delivered through the same agentic pipeline.

The trust architecture is intentional: cited sources for every recommendation, structured reasoning chains, confidence indicators on each conclusion, an audit trail of how the agents arrived at the output, and follow-up prompts that let the client challenge specific recommendations and trigger re-analysis.

How it works

Flow: client signs up → conversational onboarding chatbot identifies whether they belong to Oil & Gas or Power Utilities & Renewable Energy → routes to the corresponding department, each with its own MD Agent and engagement framework → AI-generated scoping brief proposes consulting frameworks tuned to the SME's context → multi-agent research & analysis runs in parallel → deliverable assembly with cited sources → interactive Q&A on the output → role-based deliverable views (CEO summary, CFO financial model, operations detail).

Two-tier Claude strategy: Sonnet 4.6 for MD + EM agents (strategic synthesis, narrative quality, framework reasoning); Haiku 4.5 for Senior Associates + Analysts (parallel data gathering, fast turnaround, low cost). Long-term memory enables automatic personalization across engagements, so the AI consulting team learns the client's context over time the way a real consulting relationship would.

Who it's for

Customer segment: small/medium-sized enterprises in the energy industry — oil & gas, utilities, renewable energy. End-users: decision-makers and operational leaders within energy-sector SMEs — C-suite executives (CEO, CFO, COO), strategy managers, operations managers. Most revenue-generating users: C-suite executives and senior strategy managers — they authorize purchase and consume the strategic output directly.

Engagement model: B2B, with the buyer and end-user typically the same person in an SME context (no procurement layer to navigate). Geographic positioning: energy-sector SMEs globally, with strong fit for regions where the energy transition is most actively reshaping the operating environment.

Why it matters

The global energy consulting market is $18B in 2025, growing to $24–28B by 2030 — but the bulk of that revenue concentrates at the top of the client pyramid. The long tail of energy-sector SMEs has the strategic complexity of large enterprises without the budget for traditional consulting, and AI now finally has the reasoning depth to bridge that gap with cited, auditable output.

Ai-Con's structural bet is that multi-agent architectures with role specialization beat single-monolithic-LLM approaches for strategy work — because the consulting industry already proved the pyramid structure scales reasoning under budget constraints. Same architecture, different substrate.

At a glance

Project

Ai-Con

Built by

Gilbert (Chao) Liu

One-liner

Multi-agent AI consulting platform for energy-sector SMEs — MD, EM, Senior Associates and Analysts deliver McKinsey-grade strategy work.