Shackleton AI emerges from stealth: AI agents for construction paperwork, backed by a $5M Battery Ventures seed round

Shackleton AI said on Wednesday, November 4, 2026, that it had emerged from stealth with a platform that unifies development and construction management workflows and deploys AI agents to automate administrative project work. The New York startup, founded in 2026, also announced $5 million in seed funding led by Battery Ventures, which it will use to expand the workflows its agents can manage and to accelerate its go-to-market efforts.

What the agents actually do

Shackleton brings workflows from site selection and bidding through project turnover into a single system. Its AI agents handle administrative tasks that consume project managers’ time: drafting requests for information (RFIs), following up on overdue submittals, flagging budget variances before payment requests are submitted for approval, and converting field notes into daily reports. Unlike tools that automate work inside a single application, Shackleton says its agents operate across the entire platform, coordinating activities across projects and portfolios — while people keep oversight through review and approval workflows.

The company’s product pages, as described by Unite.AI, present role-specific agents for field work, planning, development, procurement, legal work, growth, and office and finance operations, working from a shared project model of bids, budgets, RFIs, submittals, change orders, and field reports. The pitch is context: an agent that can see related records across a job, rather than only the information inside one application. Its communication hub lists Google Workspace and Microsoft 365 integrations, and its integration directory lists QuickBooks Online, QuickBooks Desktop, Sage 300 Construction and Real Estate, and Sage Intacct — a way to connect job activity to an existing accounting system instead of recreating every financial record by hand.

The problem it claims to solve

For many construction teams, project information is spread across separate systems for bidding, documents, scheduling, and finances. Shackleton says it is not uncommon for a company to juggle seven to 10 systems and for teams to spend 10 to 15 hours a week on data entry and administrative work. Tariffs, higher interest rates, and owners pushing for faster rollouts leave little room for that overhead, while legacy construction platforms were designed mainly as systems of record, leaving the follow-up to project teams.

The founders say they lived the problem. Michael Ragheb, Hassan Yasin, and Olexandr Kaltsev grew Ernest, a technology-enabled commercial general contractor, to $35 million in revenue across the Southeastern United States while struggling with fragmented software. “I was spending more time moving data between systems than running my jobs at Ernest,” said Ragheb, Shackleton’s co-founder and chief executive. He says multi-location customers have completed buildouts up to 35% ahead of schedule and as much as 18% under budget — company-reported outcomes, not a guarantee for every deployment or an independently established measure of what the AI agents contributed. Red Lobster’s senior director of construction and rebuild, Andi Cramer, said consolidating communication and project tracking on Shackleton “has kept us on track and on schedule” during the brand’s rebrand rollout. Construction software is adding agents elsewhere too: CMiC has put plain-language AI agents into its construction ERP.

The money, and what it does not prove

The $5 million round was led by Battery Ventures general partner Michael Brown, who said the Shackleton team “built this platform to run their own business before bringing it to market” and that Battery believes “their agentic approach is how construction projects will be managed going forward.” Brick & Mortar Ventures, Trust Ventures, Correlation Ventures, Motivate Partners, Thayer Investment Partners, Act One Ventures, Brainstorm Ventures, and Silence VC also participated. The announcement discloses no valuation.

What Shackleton has disclosed about traction is project value, not revenue. The platform launched in June 2026 and now manages more than $1 billion in project value across more than 15 states, serving midsize general contractors in multifamily, industrial, and commercial construction, as well as multi-location brands including Red Lobster, Dog Haus, and Pepper Lunch. That $1 billion describes construction projects managed through the software, not Shackleton’s revenue or valuation. Agent platforms launched the same day in sales, with Nooks formally launching its Revenue Agent Platform alongside a $500 million financing commitment.

Shackleton is building toward a future in which construction projects can largely manage themselves, bringing the proactive, always-on assistance people expect from consumer AI tools into project delivery. That is an ambition, not the launch claim: what launched is administrative automation with human review and approval in the loop. Whether agents that draft RFIs and flag budget variances can earn the trust to coordinate more of a job — and whether the 35% and 18% figures hold beyond the customers Shackleton cites — is what the next buildouts have to show. Together AI made a parallel agent bet the same day, with AI agents that rewrite GPU code to make open models faster overnight. Demos can be requested at shackleton.ai/demo.