Lio Cuts Contract Review Time From 120 Minutes to 2 Minutes

The Hiring Surge: Roles, Teams, and Application Process

Lio, the Munich-based AI procurement platform, has 25 open roles on its Ashby board — nearly a third of its 90-person headcount, after a $30 million Series A led by Andreessen Horowitz, TechCrunch reported. The round didn't just fund product; it funded a hiring plan that treats talent as the primary product risk.

Engineering carries six openings: Core Engineer, Forward Deployed Engineer (full-time and intern), Founders Associate in the CTO office, Head of Engineering, Security Risk & Compliance Manager, and Solution Architect. Go-to-market dominates with twelve: Account Executives in Munich and New York, Head of US Sales (New York), Product Marketing Manager, Sales Development Representatives across both offices, Sales Engineers (intern and full-time), a Value Strategist, and a Founding GTM & Branding Operator. Implementation claims five: AI Implementation Manager, Forward Deployed Engineer (intern and full-time), Founders Associate in the COO office focused on product communication and change management, and Head of Implementation Engineering. Business Operations rounds out the list with a Finance Intern and an open application slot.

Eighteen roles are full-time; seven are internships. Twenty-two sit in Munich, three in New York, a split mirroring the push to establish a U.S. sales foothold while keeping product density in Germany. Six roles are tagged explicitly on-site, though most Munich listings carry that expectation by default.

The application flow is standard Ashby: click a role, submit resume and LinkedIn, answer screening questions. No public take-homes or portfolio requirements appear at the top of the funnel. But the titles themselves signal the filter. "Forward Deployed Engineer" appears in both Engineering and Implementation, a hybrid profile that writes code at customer sites to close the gap between multi-agent demos and production procurement workflows. "Value Strategist" in go-to-market suggests a consultative sales motion, not a volume play. "Founders Associate" roles in the CTO, CEO, and COO offices show the founding team still wants direct leverage on hiring, not a delegated people function.

The rebrand from askLio to Lio, the Y Combinator pedigree, the a16z term sheet — those are the signals candidates see before they click apply. The 25 roles are the signal the company sends back: we are building what we said we would, and we need the specific people who can ship it.

Product Vision: Multi‑Agent Procurement Platform and Roadmap

Lio markets itself as the first multi‑agent system for enterprise procurement. Specialized AI agents operate in parallel across the purchasing lifecycle — researching vendors, negotiating terms, managing approvals, onboarding suppliers, matching invoices, tracking deliveries. A single request enters; within seconds, multiple agents execute end‑to‑end across ERP, P2P, email, and contract repositories. The company says this compresses RFQ cycles from weeks to days and drives up to 99 percent automation on finance workflows such as invoice processing.

The agent roster is granular: Freetext Agent, Sourcing Agent, Goods Receipt Agent, Negotiation Agent, Order Confirmation Agent, Search Agent, Approvals Agent, RFQ Agent, Supplier Onboarding Agent, Invoice Agent, Contract Negotiation Agent, Procurement Intelligence Agent, PR Review Agent, and Guided Buying Agent. Each trains on procurement best practices and the customer's organizational data. They work 24/7 with human‑on‑the‑loop controls for critical decisions. The RFQ Agent creates competition and collects quotes from multiple suppliers without buyer involvement, then delivers comparison‑ready offers with award recommendations. The Invoice Agent performs autonomous three‑way matching around the clock, comparing incoming invoices against purchase orders and goods‑received data, flagging conflicts, and attempting resolution via email, Teams, or other channels.

Lio organizes its value proposition into five layers customers can adopt in any order: Agent‑Augmented Buying, Operational Automation, Scaled Savings, Supercharge Buyers, and Agent Supervisor & AOPs (Agent Operating Procedures). The last layer introduces new internal roles — Agent Process Designer, Procurement Agent Builder, Procurement Agent Supervisor, tasked with building, managing, and optimizing agent teams. The company frames this as turning standard operating procedures into agent operating procedures, shifting procurement professionals from task execution to agent supervision.

The platform ingests unstructured data — quotations, dialogues, images, and converts it into standardized, actionable procurement requests. It enriches that data with a search interface for catalog items, vendors, and framework agreements. A virtual coach learns from procurement guidelines and guides requesters through compliant processes. Requests route automatically to the correct approver. Purchase requests validate and transform into purchase orders without human intervention, eliminating 85 percent of manual work. Suppliers upload catalogs through a dedicated portal while buyers retain approval control. The system monitors contracts continuously, surfacing only those with negotiation potential and catching hindsight savings before they expire.

Faster contract reviews (120 → 2 min) 98%, Lio's product page puts

"We treat Lio as our single source of truth, and by feeding the Lio Assistant with docs, meeting notes, and even internal podcasts, we're building an agent‑ready knowledge base that lets us scale, even as teams change." — Anna Tillmann, TÜV SÜD

Andreas Schick, COO at Schaeffler, said the collaboration "redefines the future of Purchasing at Schaeffler" and underscores "the transformative potential of agentic AI." Lio works with Fortune 500 and Global 2000 companies — airlines, manufacturers, pharma, and adoption of compliant processes reaches roughly 95 percent at enterprise customers.

Near-term roadmap milestones, signaled by the Series A and the New York office, center on global expansion and R&D depth. The company deploys AI engineers on‑site at major customers to increase time‑to‑value and deepen agent capabilities. The product vision moves toward agents that execute entire administrative workflows end‑to‑end, not just isolated tasks. Every transaction, negotiation, and outcome feeds back so the procurement AI grows smarter and more aligned with each organization's strategy over time. Implementation-to-impact targets remain weeks, not months.

Inside the Interview Bar: What Lio's Screen Actually Tests

Lio's interview process follows a two-round structure prioritizing engineering fundamentals before domain-specific depth, a candidate posted on AmbitionBox in early 2024. Round one tests general engineering aptitude, broad questions about past projects and core concepts. Round two drills into technical specialties such as electromagnetic theory and communication systems in granular detail. The reviewer described interviewers as "very nice and friendly" and rated the experience positively. A separate AmbitionBox page updated January 2025 aggregates questions across multiple Lio roles, suggesting a consistent assessment framework as headcount targets grew.

That framework aligns with how the broader AI procurement sector evaluates talent. ProcurementTactics frames its 25-question toolkit around testing how candidates think with AI, not whether they can code or build tools. The guide targets junior and mid-level procurement roles and emphasizes practical reasoning over syntax: candidates walk through how they would prompt an agent to analyze supplier risk, structure a negotiation playbook, or flag anomalous contract clauses. The rubric rewards structured thinking, awareness of hallucination risk, and the ability to design human-in-the-loop checkpoints, competencies mapping directly to Lio's product vision of such agents with those controls.

The technical bar suggests candidates need experience with unstructured enterprise data — PDFs, email threads, ERP exports, and the retrieval, chunking, and evaluation pipelines that make such data usable by autonomous agents, though no public accounts confirm final-stage evaluation criteria. Cultural fit screening is less documented. The AmbitionBox reviewer's emphasis on interviewer warmth hints at a collaborative tone, but no standardized culture-fit assessment (such as those marketed by MyCulture.ai or CriteriaCorp) has been linked to Lio's process. Resumly.ai and CriteriaCorp argue AI-driven screening can restore the humanity of hiring by offloading repetitive filtering, yet Lio's two-round model still relies on human engineers for the deep technical dive. Whether the company layers automated pre-screens or structured values assessments on top remains unverified.

What is clear: the candidate profile sits at the intersection of three vectors, strong systems engineering fundamentals, hands-on experience with LLM orchestration over messy enterprise data, and a procurement-domain mindset that treats compliance and auditability as first-class constraints. Those roles, but the technical bar described applies most directly to the engineering cohort. Product and GTM candidates likely face a parallel track swapping electromagnetic theory for go-to-market strategy exercises and procurement workflow mapping, though no first-hand accounts of those tracks have surfaced.

Market Impact: How Lio's Hiring Signals Growth in AI Procurement

Lio's 25-role push lands as the AI procurement market separates into two tiers: incumbents bolting generative features onto legacy suites, and AI-native platforms building multi-agent architectures from the ground up. Three independent forecasts bracket the 2025 market between $3.15 billion and $6.17 billion; by 2034–2035 they project $14.6 billion to $39.2 billion, compound annual growth rates between 18.6% and 28%. The U.S. segment alone, valued at $1.12 billion in 2025, is on track for $13.5 billion by 2035 at a 28.2% CAGR. North America commands roughly 40–45% of global revenue; Asia Pacific is forecast to grow fastest through 2034.

IDC's figures put global enterprise AI spending at $184 billion for 2026. McKinsey's 2026 survey found 65% generative AI adoption across organizations and 63% reporting positive ROI. In procurement specifically, 94% of executives now use generative AI weekly, McKinsey's 2026 survey found, and 80% of chief procurement officers rank AI investment as a top priority. That demand pull explains why software captures 64–70% of market revenue, cloud deployment holds 72% share, and machine learning leads technology segments at 43%. Spend analytics remains the largest deployed application, though supplier management and risk analytics are gaining share.

Incumbents are reacting. Coupa added predictive procurement, supplier risk analysis, and automated sourcing to its Autonomous Spend Management platform in May 2025. SAP followed in March 2025 with generative AI for supplier search, contract analysis, and guided sourcing inside Ariba. SAP's acquisition of Coupa consolidated two of the largest rule-based suites under one AI roadmap. Meanwhile, emerging specialists raise growth capital: Omnea closed $50 million for automated supplier assessment and performance analysis. Arkestro, Fairmarkit, and Sievo are cited across reports as AI-native sourcing and analytics platforms gaining enterprise traction. GEP, Ivalua, Jaggaer, Zycus, Globality, Basware, and SynerTrade round out the competitive set.

Lio's multi-agent approach — autonomous agents executing end-to-end sourcing workflows from spend analysis through contract negotiation and purchase-order placement, sits at the frontier these forecasts describe. TrendX Insights identifies autonomous procurement agents, generative category strategy generation, real-time market intelligence, and carbon-integrated scoring as the next capability wave. The services segment is projected to grow at the highest CAGR (24.5%), NLP is the fastest-growing technology segment, and risk management with predictive analytics leads application growth. SMEs are expected to outpace large enterprises in adoption rate through 2034.

Against that backdrop, a 25-person hiring plan concentrated in those areas is a capital-allocation signal. It says Lio bets on product velocity to win enterprise contracts before incumbents retrofit their installed base and before better-funded AI-native rivals lock up the same talent pool. The market's trajectory — $184 billion in enterprise AI spend, 94% weekly generative AI usage among procurement leaders, a shift from rule-based automation to autonomous agents, means the window to establish a defensible multi-agent product is measured in quarters, not years. Lio's screen for specialized engineering and product talent is effectively a filter for the skill set the entire sector now competes for.

Reactions: Talent Pool, Competitors, and Investors Respond

Lio's $30 million Series A, announced via PRNewswire and backed by Andreessen Horowitz, has functioned as a market signal as much as a capital event. The round, closed while the company was still askLio, arrived alongside a rebrand and a public claim of more than 100 enterprise clients, including Fortune 500 accounts, managing billions in spend. That combination of fresh capital, named customers, and a multi-agent architecture claim has pulled three audiences into motion: candidates, rivals, and investors watching the category form.

Talent Pool: Applications Follow the Signal

The broader technology labor market is tight. Robert Half's second-half 2026 survey of technology leaders found 78% plan to increase permanent headcount, up from 61% earlier in the year, and 66% expect to add contract professionals. At the same time, 65% of hiring managers say finding skilled talent is more challenging than a year ago, and 71% report project delays tied to skills shortages. AI and ML roles specifically posted 49,200 openings in 2025, a 163% year-over-year jump, Robert Half found. Against that backdrop, a Series A backed by a16z, with a public narrative about collapsing weeks of work into minutes and replacing outsourced labor with AI, acts as a beacon for engineers and product people who want to work on agentic systems that have already cleared the proof-of-concept hurdle.

Stanford AI Index's data shows U.S. private AI investment hit $285.9 billion in 2025, yet the flow of AI researchers into the United States has dropped 89% since 2017, Stanford AI Index reported, with an 80% decline in the last year alone, Stanford AI Index's data shows. That supply constraint makes any company demonstrating traction, Lio cites an 85% reduction in manual work for procurement buyers and 10% additional procurement savings, disproportionately attractive to candidates who can afford to be selective. The company's 25 roles span those areas, signaling a move from core R&D into commercial scaling. For a candidate, that means the technical risk has shifted from "can the agents reason?" to "can we ship, sell, and support this at enterprise grade?", a different and often more appealing problem set.

Competitors: The Category Is Crowding

Lio does not operate in isolation. CB Insights and startuphub.ai both list Omnea as a direct alternative, describing its platform as covering request submissions, approval workflows, purchase order creation, vendor onboarding, and supplier risk management. Procurement Magazine's "Top 10 Procurement Start-Ups to Watch" includes Lio and highlights its coordinated network of specialized agents managing the full lifecycle, researching vendors, negotiating terms, verifying compliance, tracking deliveries. The same list implicitly frames the competitive set: companies trying to automate the same end-to-end process, often with single-agent or workflow-automation approaches rather than the multi-agent parallelism Lio emphasizes.

The competitive pressure is visible in funding velocity. Private AI investment more than doubled in 2025, with generative AI growing over 200% and capturing nearly half of all private AI funding. Newly funded AI companies rose 71%, and billion-dollar funding events nearly doubled. In procurement specifically, the shift from robotic process automation to agentic AI is the current battle line. Omnea and others race to add reasoning layers atop their workflow engines; Lio's bet is that it started with the reasoning layer. The hiring push, especially in go-to-market roles, suggests Lio believes the window to define the category's buying criteria is open now, and that sales capacity is as strategic as engineering velocity.

Investors: a16z's Thesis on Display

Andreessen Horowitz's public write-up, "Investing in Lio," frames the company as replacing outsourced labor with AI, a thesis aligning with the firm's broader view that agentic systems will absorb services revenue, not just software budgets. The a16z piece cites the same 85% manual-work reduction and 10% savings figures, and notes "more than 100 clients globally, including Fortune 500 companies across industries." That specificity in a venture blog post is unusual; it reads as a signal to follow-on investors that the metrics are auditable and the reference customers are real.

The broader investor reaction can be inferred from category momentum. Deloitte's 2025 State of AI in the Enterprise survey found only one in five companies has a mature governance model for autonomous AI agents, yet agentic AI usage is poised to rise sharply in the next two years. That gap — demand outpacing governance, is exactly where procurement agents live: they touch spend, compliance, and vendor data, all of which require audit trails. Investors watching Lio are effectively betting on whether a multi-agent architecture can satisfy enterprise governance requirements faster than incumbents can bolt agents onto existing suites.

The Stanford AI Index adds context: industry produced over 90% of notable frontier models in 2025, and on SWE-bench Verified, performance rose from 60% to near 100% in a single year. The code-generation capability underpinning agentic tool use is no longer the bottleneck; the bottleneck is integration, trust, and go-to-market execution. Lio's 25-role hire — heavy on product and GTM alongside engineering, is a capital-allocation decision saying the technology risk has retired enough to spend Series A dollars on distribution. Other investors will treat that allocation as a data point: if a16z is comfortable funding sales heads at this stage, the bar for the next round has moved from "demo works" to "revenue scales."

The Feedback Loop

The three audiences reinforce each other. Candidates see a funded competitor set and a category leader claiming hard ROI numbers; they apply. Competitors see the talent inflow and the a16z endorsement; they accelerate their own hiring and messaging. Investors see the talent war and the competitive density; they price the next rounds accordingly. Lio's hiring push is not just a response to its own roadmap, it is a move in a market-wide game where the cost of waiting is measured in engineers who sign offer letters elsewhere, enterprise buyers who standardize on a rival's platform, and term sheets that go to the company that proves it can hire faster than the category can commoditize.

From askLio to Lio: A Brief History of the Company's Evolution

Lio's current hiring push, 25 open roles across engineering, product, and go-to-market, didn't emerge from a vacuum. It traces back to a founding insight that procurement, not AI, was the real problem to solve.

The company began in 2023 in Munich under the name askLio. Three founders — Vlad Keil, Lukas Heinzmann, and Till Wagner, launched it after Keil lived the procurement bottleneck twice: first as an employee inside a large enterprise, then again while selling enterprise software at an earlier startup. "When we were selling enterprise software, we had to go through procurement ourselves and saw how manual and fragmented the process still is," Keil told TechCrunch in March 2026. The frustration wasn't abstract. Employees at large companies drown in forms, approval chains stretching weeks, and thousands of free-text requests that existing systems can't parse automatically. Procurement teams process them manually because they have no other choice.

That insight carried the trio into Y Combinator's Summer 2023 batch, still as askLio. The accelerator gave them a laboratory to test whether an AI system could ingest natural-language requests, structure them, source suppliers, negotiate terms, and execute purchases end-to-end, across ERPs, inboxes, contracts, and the open web. The answer came fast. By the time the company rebranded to Lio (the exact timing isn't public, but the new name appears on the Series A announcement), its agents were managing billions in enterprise spend for dozens of Global 2000 and Fortune 500 customers, Munich Re, Brose, Novozymes, and enterprises across chemicals, postal services, retail, transportation, medtech, and pharmaceuticals. The platform now serves more than 100 global enterprises, is ISO 27001 certified, EU-GDPR compliant, and deploys in under two weeks.

The rebrand signaled a shift from "ask", a consumer-grade interface, to "Lio": a virtual procurement workforce. The distinction matters. Legacy vendors like SAP Ariba and Oracle built software to help humans do procurement faster. Lio builds agents that execute the workflow themselves. Keil calls this Agent Operating Procedures (AOPs): standard operating procedures encoded so AI agents can triage requests, analyze quotes, compare suppliers, negotiate, onboard vendors, and complete transactions without human handoffs at every step.

Investors noticed. On March 5, 2026, Lio announced the Series A, with participation from SV Angels, Harry Stebbings, and Y Combinator, bringing total funding to $33 million. a16z partner Seema Amble framed the bet: "We're entering a phase in the enterprise where AI moves beyond workflow co-pilots to autonomous, multi-agent execution." The capital is earmarked for U.S. expansion and deeper agent capabilities, precisely the work the 25 new hires will advance.

Today the team sits at roughly 80 people, per Y Combinator data, with a presence in Germany. The founding trio's original frustration — procurement as a language problem between people and rigid systems, now sits at the center of a category shift. Enterprises spend over $180 billion annually on procurement talent versus roughly $10 billion on procurement software, a gap that exists because the work never fully automated. Lio's hiring surge is the direct consequence of proving that gap can close. The 25 roles aren't growth for growth's sake; they're the next layer of infrastructure needed to turn early enterprise wins into a repeatable, scalable platform that can onboard the next hundred customers without the founders in every room.

What This Story Does Not Cover: Scope Limits and Out‑of‑Topics

This article tracks Lio's 25-role hiring push across engineering, product, and go-to-market functions and what that push signals about the maturation of multi-agent procurement software. It does not pretend to be a compensation survey, a sector-wide AI landscape, or a financial audit. Drawing those lines keeps the piece useful instead of sprawling.

Salary specifics for Lio's open roles are not here. The research surface on AI engineer pay is thick — levels.fyi reports median U.S. AI engineer compensation at $245,000, with staff-level specialists commanding an 18.7% premium over non-AI peers as of 2025, but none of those figures come from Lio itself. The company has not published band data for its 25 openings, and Zero G Talent's board shows listings for other employers (ASML at $21k–$356k median $154k; Stripe at $25k–$336k median $235k) without a Lio equivalent. Readers who need offer-level numbers should treat this piece as a signal to ask recruiters directly, not as a substitute for that conversation.

Unrelated AI sectors fall outside the frame. The generative AI boom spans foundation model labs, coding assistants, video generation, drug discovery, and more. Lio sits in a narrow wedge: multi-agent systems that execute procurement workflows, vendor research, negotiation, approvals, delivery tracking, in parallel. The a16z backing notes 100-plus global clients including Fortune 500 companies managing billions in spend, with 85% reduction in manual work and 10% additional savings cited. That traction is procurement-specific. Conflating it with the broader AI hiring surge would blur the actual hiring signal: Lio needs engineers who understand agent orchestration, procurement domain logic, and enterprise integration, not just transformer architecture.

Deep financial audits are excluded. The only audited financial statement in the research belongs to Lion Energy Limited, an Australian entity with a December 2025 filing, a different company entirely. Lio's private funding history (a16z participation, prior rounds) is public in outline but not in granular cap-table or runway detail. This article does not reconstruct burn rate, runway, or valuation markup. Those questions belong in a diligence memo, not a hiring trend story.

What remains in scope: the composition of the 25 roles, the product milestones those hires are expected to hit, the interview bar Lio applies, and how competitors and candidates are reacting. Each of those threads appears in earlier sections. This section exists so readers don't mistake the article for something it never promised to be.

The signal is in the roles; the noise is everything this section names and sets aside. When Vlad Keil walked through a procurement department years ago, he saw the same forms, the same approval chains, the same free-text chaos that Lio's agents now parse in seconds. The 25 open roles are the proof that the fix works — and that the next hundred enterprises are waiting.

Working in frontier tech? Zero G Talent tracks the openings: see every open ASML role, browse frontier tech jobs, openings at Stripe, and the people building the field.