Logistics · Deep dive

Augment

San Francisco AI startup from Deliverr co-founder Harish Abbott whose freight-native 'AI teammate,' Augie, reads email, makes phone calls, logs into TMS platforms and carrier portals to run broker and carrier workflows from order to cash — $110M raised in five months of 2025, $35B in freight under management by September 2025, Penske Logistics deployed in January 2026, and an April 2026 acquisition pushing it into $8T wholesale distribution.

emerging

The question that decides it: Augie's wedge is executing the work no single system owns — phone calls, emails, portal logins, document chasing stitched across other companies' TMS platforms — sold as a horizontal teammate rather than a system of record. Does that layer keep its seat once the platforms that own the transaction data bundle native agents (Turvo, today Augment's integration partner, McLeod, and the largest brokers building in-house — C.H. Robinson was already automating ~75% of LTL orders itself by 2025), or does the April 2026 all-equity Merlin acquisition into $8T wholesale distribution concede that the freight back-office alone is too shallow a prize to carry $110M of venture capital?

Snapshot

Augment is what Harish Abbott did with the conviction he carried out of selling Deliverr to Shopify for $2.1 billion: that the real bottleneck in American freight is not trucks or software but the human hours burned on emails, check calls, quotes, appointment scheduling and invoice chasing that no TMS ever absorbed. Its product, Augie, is a freight-native AI agent that reads, writes, calls, listens and logs into systems, running broker and carrier workflows from order to cash. The company emerged from stealth in March 2025 with a $25 million 8VC seed, raised an $85 million Redpoint-led Series A five months later — $110 million total by September 2025, with no disclosed valuation — claimed $35 billion in freight under management at that date, converted a Penske Logistics pilot into a production rollout in January 2026, and bought its way into the $8 trillion wholesale-distribution market in April 2026. It matters because it is the largest-capitalized bet that the freight back office gets automated by a horizontal agent layer rather than by the systems of record themselves.

Founding story

Abbott is on his third logistics company and second act with essentially the same team. After early work in Amazon fulfillment and co-founding Symphony Commerce, he and Michael Krakaris founded Deliverr in 2017 to give small ecommerce merchants Amazon-grade two-day delivery through an asset-light fulfillment network. Deliverr raised roughly $491 million — capped by a $240 million Tiger Global Series F at a $2 billion post-money (2021) — and sold to Shopify in May 2022 for $2.1 billion, the largest acquisition in Shopify’s history. Abbott became a Shopify VP; Shopify later offloaded the logistics arm to Flexport in June 2023, a quiet epitaph for the strategy that had justified the price.

What Abbott says he took from Deliverr was a specific frustration: Deliverr moved goods but never fixed the coordination layer underneath — the operational traffic that consumes the majority of human time in freight (Business Insider, March 2025). In 2024 he founded Augment to attack that layer directly, and the team is the tell: co-founder and CTO Artur Rivilis had been with Abbott at Symphony Commerce, led all product and engineering at Deliverr, then was a Shopify engineering VP with a Flexport stint; co-founder and CCO Justin Hall brought 3PL operating experience. Glassdoor reviewers note that almost all early employees came from the founders’ previous startup (2025-26). Augment is less a new company than the Deliverr organization pointed at a different problem, with the founder’s own credibility — and Shopify Ventures’ money — attached.

How it works

Augie is best understood as a synthetic freight operator with logins. A brokerage connects it to its TMS (Turvo is a named native integration, announced late 2025; the company also cites ERP, email, voice and messaging connections), grants it a mailbox and a phone number, and configures workflows modeled on the customer’s own standard operating procedures. Then the work arrives the way freight work actually arrives: unstructured. A shipper emails three loads in a PDF; Augie extracts the details and builds the loads in the TMS. A quote request comes in; Augie prices against contract rates and drafts the response. A truck goes quiet; Augie calls the driver or dispatcher, transcribes the answer, updates the shipment record, and escalates to a human only when something is genuinely off — a rejected appointment, a rate dispute, a detention claim. At delivery it chases the POD and lumper receipts across carrier portals, assembles the document packet, and pushes billing.

Two design choices distinguish it from the RPA bots and single-task voice agents that preceded it. First, shipment-level ownership: Augie holds the context of an entire load across channels rather than executing one scripted task, which is what lets Armstrong Transport Group reps go from managing 10 loads a day to 20-30 (company case study, 2025). Second, accumulated context: the February 2026 Knowledge Hub turns each customer’s SOPs, tribal knowledge and operating history into a queryable layer — a ChatGPT trained on how this brokerage runs (FreightWaves, February 2026) — which deepens switching costs with every month of use. That accumulation is the intended moat; whether it is one is the open question below.

Product and business overview

The platform now has three named layers. Augie, the flagship agent, sold into four segments — freight brokerages, FTL carriers, LTL and last-mile carriers, and (since April 2026) B2B distribution — with productized workflows for load building, carrier selection, track and trace, appointment scheduling, document management and customer service. Workflow Builder, the configuration surface where customers encode their SOPs, positioned as natural-language setup rather than custom code. Knowledge Hub (launched February 9, 2026), the retrieval layer over a customer’s operational knowledge, born — per Abbott — from Augie users asking operational questions mid-shift (Supply Chain Matters, February 2026). The Merlin acquisition (April 29, 2026, all-equity) extends the same agent into wholesale distribution quote-to-cash — supplier portals, order entry, purchasing — under Alex Moazed, the Applico founder who joined as President of Wholesale Distribution; Abbott told Axios more early-stage AI acquisitions are coming (April 2026). The enterprise posture is deliberate: SOC 1 and SOC 2 Type II certification, SSO enforcement and third-party pen testing headline the site — table stakes for selling to Penske and NFI, and a real barrier for smaller rivals.

Business model and pricing

Augment publishes no pricing; every path on its site terminates at a demo request, and third-party directories confirm pricing is contact-only (AI Agent Store, 2025). The company describes itself as an “AI productivity platform,” and the sales motion is classic enterprise: multi-month pilots converting to production contracts, as with Penske’s six-month pilot before the January 2026 rollout. What the company markets instead of price points is ROI arithmetic: Armstrong Transport cut invoice delays 40%, accelerated billing by eight days and recovered 5%+ gross margin per load (company, 2025); Transportation One projected seven figures in annualized savings (September 2025); Penske anticipates 30-40% productivity gains (January 2026). The unpublished economics cut both ways. Value-based enterprise contracts against numbers like those support real ACVs, and “$35B in freight under management” is a deliberately revenue-shaped metric that isn’t revenue — Augment takes no freight margin, unlike a digital broker. The absence of any disclosed ARR, seat count or per-load rate through July 2026 means the company’s actual revenue scale is unverifiable from the outside, and rivals like Pallet price per completed task specifically to attack platform-sized contracts.

Traction over time

The trajectory is unusually fast even by 2025 AI standards — $110 million inside six public months, and the September 2025 Series A came one day after HappyRobot’s Series B, a signal of how hot the category ran. The conspicuous gap: the $35 billion figure has not been publicly updated since September 2025, and no revenue, retention or customer-count time series exists (through July 2026).

Market analysis

The labor pool Augment is automating is enormous relative to the software market that currently serves it. US freight brokerage alone was a ~$58-66 billion revenue industry in 2025 (SkyQuest; Fortune Business Insights), with labor near 40% of broker cost structures, much of it the repetitive communication Augie targets. The software wrapper is far smaller — freight broker software estimated at ~$2.7 billion in 2025 (WiseGuyReports) — but the AI-native segment is being sized separately and aggressively: one 2025 estimate put AI-powered freight back-office automation at $2.3 billion growing ~31% annually to $25.6 billion by 2034 (MarketIntelo). The bull structural case is that agents are priced against labor, not software, so the addressable spend is the payroll line; the Merlin deal extends the same logic to wholesale distribution, an industry Augment values at over $8 trillion in US throughput (April 2026). The bear structural force is the 2023-26 freight recession: brokers cutting headcount adopt automation to survive, but their software budgets shrink with their margins, and a February 2026 sell-off in traditional brokerage stocks driven by AI-displacement fears (MarketMinute) cuts both ways — validation of the thesis, and a warning that Augment’s customer base itself may be the layer AI compresses.

Competitive intel

The competitive set (scale and angles in frontmatter) splits into three rings. The funded startups — HappyRobot (~$500M valuation, September 2025), Vooma, FleetWorks, Pallet, plus a long tail of CloneOps, LunaPath and Freight Hero — mostly entered through voice or single workflows; Augment’s differentiation is scope (whole-shipment ownership, order to cash) and capital, holding roughly 2x the funding of its nearest direct rival. The incumbentware ring is more dangerous: every TMS vendor watching agents operate on top of its records — including Turvo, currently Augment’s partner — can bundle a native agent, and the system of record holds the data, the contract and the renewal. The third ring is in-house AI at the scaled brokers: C.H. Robinson was already automating ~75% of LTL orders with proprietary systems handling 10,000+ daily messages by 2025, which suggests the very largest accounts will build, not buy. Augment’s counter-positioning is the Deliverr network — founder-led distribution into enterprise shippers and 3PLs that pure-play AI startups can’t call on — and the bet that cross-system, cross-channel execution is precisely what no single TMS can own. Penske choosing Augment over building is the best current evidence for that bet (January 2026).

History and evolution

• May 2022 — Shopify acquires Deliverr for $2.1B; Abbott and Rivilis become Shopify VPs.
• Jun 2023 — Flexport buys Shopify Logistics including Deliverr; Rivilis passes through Flexport.
• 2024 — Abbott, Rivilis and Hall found Augment in San Francisco; builds in stealth with 8VC.
• Mar 18, 2025 — Emerges from stealth with $25M seed led by 8VC; launches Augie for FTL, LTL, expedited and drayage workflows.
• Sep 4, 2025 — $85M Series A led by Redpoint ($110M total); claims $35B freight under management, dozens of enterprise customers; announces 60+ planned hires. Lands one day after HappyRobot’s $44M Series B.
• Late 2025 — Turvo partnership: Augie embedded in Turvo’s TMS for order-to-cash orchestration.
• Dec 9, 2025 — Abbott on Bloomberg’s Talking Transports — mainstream analyst attention.
• Jan 14, 2026 — Penske Logistics converts six-month pilot to production rollout; ~600K loads validated, 30-40% productivity target.
• Feb 9, 2026 — Knowledge Hub launches; NFI an early named user.
• Apr 29, 2026 — Acquires stealth startup Merlin (all-equity); enters wholesale distribution; Alex Moazed joins as President; Abbott signals more acquisitions (Axios).
• Jul 17, 2026 — Named AI Solution Provider winner, FreightWaves 2026 AI Excellence awards.

What people say

The case for. Customer proof points are unusually specific for a company this young: Armstrong Transport’s doubling-to-tripling of loads per rep with recovered margin (company case study, 2025), Transportation One’s seven-figure projected savings (September 2025), and Penske — a name that does not adopt vendor tech casually — publicly crediting the pilot and citing NPS among its own associates as a rollout driver (PR Newswire, January 2026). Trade press treats Augment as a category leader: FreightWaves gave it a 2026 AI Excellence award (July 2026), and the Journal of Commerce framed the back-to-back Augment/HappyRobot raises as the moment AI-in-logistics went mainstream (September 2025). Glassdoor reviews (small base, 2025-26) praise a brilliant, kind team, competitive pay and real shipping velocity.

The complaints. No G2 or Capterra review base exists yet (July 2026) — itself a data point for an enterprise vendor 16 months post-launch, consistent with a small number of large accounts. The recurring critical themes come from adjacent evidence. Industry practitioners report edge cases that still break freight voice AI: heavy accents, loading-dock noise, three-way calls, and disputes requiring contract interpretation (Numeo carrier guide, March 2026), and some brokers hang up on detected AI voices. Abbott himself has publicly conceded the category’s uncomfortable truth — pilots succeed and production deployments often don’t, because of change management, integration politics and trust gaps (FreightCaviar, June 2026); candid, but a description of his own funnel risk. Glassdoor’s noted downside is that nearly all employees came from the founders’ prior startup, making outsiders feel like late arrivals (2025-26). And the skeptic’s read of the disclosure pattern writes itself: no valuation, no revenue, a headline metric ($35B FUM) that measures customers’ freight rather than Augment’s income and hasn’t been refreshed since September 2025, and a pivot into a second $8 trillion industry twelve months after launching into the first.

Outlook: the open question

Augment works if, by roughly the end of 2027, it can show that the horizontal layer compounds: Penske-class logos renewing and expanding after their pilots, Knowledge Hub making Augie measurably harder to rip out than a voice bot, disclosed revenue traction that justifies the next round at a step-up, and TMS platforms continuing to partner (as Turvo did in late 2025) rather than bundle competing agents — proving that cross-system execution is a durable product line, not a feature waiting to be absorbed. The bull case is genuinely strong: the best founder-market fit in the category, 2x the capital of any direct rival, the only public enterprise-fleet production deployment (Penske, January 2026), and pricing against a labor line that dwarfs the software TAM. Augment fails if the systems of record close the gap first — Turvo or McLeod shipping native agents into the transaction data they already own, C.H. Robinson-style in-house builds capping the top of the market while Vooma and Pallet undercut the bottom on price — leaving Augie squeezed into mid-market brokers whose freight-recession budgets can’t support enterprise contracts; in that world the Merlin deal reads not as ambition but as an early admission that freight alone couldn’t carry the valuation, and the endgame is a sale to the TMS or broker that needed the team. The tells to watch: whether the $35 billion FUM number gets a public update with a revenue figure attached, whether the next funding round discloses a valuation, whether Turvo stays a partner through 2027, and whether any customer publicly expands Augie from freight into the Merlin-built distribution workflows — the first real test of whether this is a platform or two products sharing a brand.

Sources and further reading

• AI logistics startup Augment, from Deliverr’s founder, raises massive $85M Series A (TechCrunch, September 4, 2025)
• Augment Raises $85M Series A to Supercharge Augie, The AI Teammate for Logistics (Business Wire, September 4, 2025)
• He sold Deliverr to Shopify for $2.1 billion. Now his new startup is betting big on an AI assistant named Augie. (Business Insider, March 18, 2025)
• Deliverr co-founder launches ‘AI teammate’ for logistics, raises $25M (FreightWaves, March 2025)
• Penske Logistics Accelerates Agentic AI Supply Chain Initiative with Augment (PR Newswire, January 14, 2026)
• Augment brings institutional knowledge into the flow of freight (FreightWaves, February 9, 2026)
• Exclusive: Augment buys Merlin to deepen supply chain AI expertise (Axios Pro, April 29, 2026)
• Augment Expands Into Wholesale Distribution With Merlin Deal (Transport Topics, May 5, 2026)
• The AI Pilot Worked…Now What? (FreightCaviar, June 5, 2026)
• HappyRobot raises $44 million to expand AI agents for freight operators (Reuters via US News, September 3, 2025)
• Startups land collective $129 million in latest AI push into logistics (Journal of Commerce, September 5, 2025)
• How Armstrong Transport Group Doubled Productivity with Augie (Augment case study, 2025)