AudaciousVertical AI for the $1 trillion home-services industry.
Avoca is the AI front office built for service businesses: it answers every call, works every lead, and checks in on every customer, keeping boards full and teams focused on the work that matters. Proudly built for the trades, and trusted by 800+ operators across HVAC, plumbing, electrical, roofing, pest control, automotive, and more. Under the hood, real-time voice AI holds live conversations on noisy job sites and routes calls into legacy CRMs like ServiceTitan to book the job, the hard part isn’t the LLM, it’s everything around it.
Apurva Shrivastava
MIT CS, second-time founder. Scaled Retool as an AI engineer; built Avoca’s original voice product.
Tyson Chen
MIT CS, ex-PM at Nuro, ex-BCG. Owns GTM and the services-industry domain.
Traction, funding & whether the hype is real.
Production, not a demo reel
Avoca 10×’d revenue in 2025, and more than $1B in jobs has been booked through the platform, production software running revenue-critical workflows.
$125M+ Series B
Raised at a $1B valuation. Backed by Audacious, with General Catalyst and Meritech leading and Kleiner Perkins, Amplify, Nexus, and Y Combinator participating.
$1 trillion, early-2000s software
Strongest in HVAC today, expanding into restoration, towing, auto, plumbing, and electrical, each one where HVAC was two years ago.
The product & the engineering underneath.
An end-to-end AI platform that turns customer demand into booked jobs: answers inbound calls 24/7 with voice AI, runs SMS and email follow-ups, qualifies intent, checks technician availability, books and reschedules jobs, coordinates dispatch between customers, CSRs and techs, and feeds analytics and coaching back into the loop.
Systems problems, not research problems.
Live calls from job sites and truck cabs, power tools, road noise, multiple speakers, accents, mid-sentence corrections, reasoned through in low latency while holding conversational state.
Scheduling, rescheduling, dispatch, and technician routing all changing mid-call while the agent is still talking.
Full read/write into ServiceTitan and other systems of record built two decades ago. Going deep here is the work.
One missed edge case isn’t a dashboard blip, it’s a $15K HVAC job that didn’t get booked. Code to customer revenue, no abstraction layers.
Stack: Node.js, TypeScript, and React, with Cursor in the workflow and a proprietary mix of foundation models. No in-house ML research org, the work is production systems engineering: latency, state, audio robustness, legacy integrations, and multi-tenant enterprise scale.
How the team works.
Apurva, directly. No head-of-engineering layer between senior ICs and the CEO, no VP of Engineering, no managers. Every engineer reports to Apurva.
Zero engineering turnover. Everyone’s there voluntarily, and the numbers back up the culture.
Recent coverage.
Avoca raises $125M+ at a $1B valuation
To power America’s services economy with AI. Read →
A chance encounter in Texas sparked a $1B AI startup
Fortune on Avoca’s rise with Kleiner Perkins. Read →
They met at an MIT poker night
Now their $1B startup sells AI to HVAC companies, dubbing technicians “main characters.” Read →
AI vs human: call handling in home services
Avoca on the To The Point home-services podcast. Watch →
Why Join: Avoca
A GTM deep dive on the problem, market, product, and team, and why Avoca keeps landing at the top of candidates’ lists. Read →
What they’re hiring for.
Five open roles on Avoca’s engineering team, New York HQ, in-office. Forward Deployed Engineering is where the team is growing fastest.
Stack: TypeScript, React, Next.js, Vercel, AWS, Supabase and Postgres, ClickHouse, Twilio and Dialpad for telephony, Datadog, with Anthropic and OpenAI models under the voice layer. What matters most is strong fundamentals and real problem-solving. Use AI tools all you want, as long as you can debug and reason about what ships.
Staff Software Engineer, Product
- Own architecture and end-to-end delivery of the hardest, highest-stakes systems
- Set the bar for code quality, testing, and reliability across the org
- Partner with Product, CS, and GTM leadership on roadmap and strategy
Senior Software Engineer, Product
- Own features end to end, backend services to AI-powered UX
- Turn customer feedback into rapid iteration and experimentation
- Work with Product, CS, and GTM; 4-6 yrs full-stack
Forward Deployed Engineer
- Roughly half the job is building and testing agents; the rest is core engineering and customer work
- Lead enterprise pilots and multi-brand rollouts, including onsite with operators
- Recent builds: self-serve controls for AI tool-call responses, custom IVR agents, on-prem CRM integrations
- The face of Avoca to CTOs, VPs of customer experience, marketing, and data teams
- Custom work ships behind feature flags; what proves broadly useful graduates into the core product
Agent Product Manager
- Own one core product line from discovery through deployment
- Go deep on one area, not a generalist PM
- Connect customer reality to engineering; synthesize CSM and TAM feedback
IT Manager
- First IT hire; own the full tech environment and SaaS stack
- Lead the IT build-out of the NYC HQ as the team scales 100 to 200+
- Own endpoint security, identity and access, and IT policy
Build AI that actually runs businesses.
In office · NYC. If you want to ship production systems where code maps straight to revenue, this is the work.