Building an Autonomous Enterprise for Real-World Services with Netic Founder Melisa Tokmak
In a Nutshell
Netic founder Melisa Tokmak built an AI platform that autonomously handles customer interactions and operations for essential service businesses like HVAC, plumbing, and roofing—managing everything from initial calls to technician deployment for billion-dollar companies. The system now serves as the primary customer interface for over 70% of clients, generating $600M+ in revenue through AI-driven interactions while freeing companies to focus solely on labor quality and service delivery. This represents a shift from traditional cost-cutting AI applications to revenue-generating automation in traditionally underserved industries.
These notes were generated by AI and may contain inaccuracies.
Melissa Tokmak is the founder and CEO of Netic, a company that builds AI for real-world services including HVAC, pet care, roofers, and other essential service businesses. Prior to founding Netic, Melissa served as Director of Engineering at Scale.AI working on go-to-market for various aspects, and previously worked at Meta.
Netic builds AI to run millions of real-world businesses that keep the world running. The company works with large enterprises in essential services, including billion-dollar revenue home services companies in HVAC, plumbing, and electric, as well as consumer wellness companies, hospitality, automotive, and pet services.
Essential services encompass every single thing needed to run life or desired activities. These businesses are typically quite large and interact with millions of end users, either consumers or other businesses. Netic operates between the company and its customers, handling customer needs and wants, matching them with operational rules, and deploying services or labor.
When a customer calls an HVAC provider, Netic agents handle the interaction across any medium—phone calls, text messages, or website contact. The system manages voice calls and provides the actual voice interface. Netic agents gather information about the customer's home, check existing records, and handle the reasoning for deploying appropriate service personnel.
The operational needs are very complex. Understanding customer needs involves determining unit types, specific requirements, service timing, and customer lifetime value. The system must decide whether to deploy the best specialist for particular systems like boilers or new-age systems, optimizing for customer delight and revenue generation.
Before Netic, these processes were handled entirely by people. These companies are extremely large but cannot grow because they must continuously invest in personnel. Many are EBITDA businesses owned by private equity, requiring focus on margins. The most important labor is the blue-collar workers delivering the actual jobs. Companies invest in customer-facing labor because these are commodity businesses where unanswered calls result in lost customers to competitors.
A typical day at a billion-dollar revenue company might start at 4:00 or 5:00 AM with no staff present. Technicians begin at 6:00 AM, but managers may find multiple employees have quit or failed to show up. Customer calls begin piling in during events like heat waves. These are cyclical businesses requiring value capture during peak seasons—winter, summer, holidays—to sustain operations year-round.
While many customers initially approached Netic for overflow call handling, over 70% are now AI-first or "Netic-first," with all customer first interactions handled by Netic agents.
Three reasons drove the decision to build rather than pursue roll-ups:
- The founder's skill set from Scale.AI focused on building real-world businesses rather than M&A
- Personal identity as a builder and engineer rather than an M&A professional
- Product scale limitations in roll-ups where products serve only acquired companies rather than being broadly applicable
At Scale.AI, Melissa built the government business unit and large enterprise units across logistics, manufacturing, financial services, and healthcare. She left before the Meta licensing deal valued around $28-30 billion.
Melissa grew up in a small town in Turkey with limited resources, arriving in the US on a full Stanford scholarship without previously owning a computer. Her background connects her to the industries Netic serves. The mission combines tangible real-world impact with solving the unsolved problem of using AI in mission-critical workflows for fully autonomous execution.
While the infrastructure work at Scale.AI was rewarding, it lacked the product edge Melissa sought—building a product that could scale and compound over time.
The vision is enabling every real-world business to run on Netic, allowing companies to focus on their core competencies—labor quality and service differentiation—while Netic handles everything else.
While robotics represents a future chapter in AI for the real world, it remains far off for essential service industries. Current buildings would require 3D printing or standardization for robotics deployment. Robotics capabilities today lack sufficient dexterity for handling varied screws, navigating different home types, accessing confined spaces, or dealing with unclear repair needs requiring wall demolition. These industries also involve significant human elements—customers often experiencing their worst days (flooded homes) or seeking stress relief (tennis at Bay Club).
The leading AI labs are not viewed as competitive risks but rather as potential partners. Key differentiators include:
- Enterprise focus on stability versus labs' rapid product changes
- Enterprise preference for proven reliability over cutting-edge but unstable solutions
- Need for focus on specific vertical problems rather than generalizable AGI approaches
- Requirement for last-mile solutions involving harnesses, orchestration, software, and product layers beyond base models
Companies like Netic must excel at all three layers: base models, harnesses and orchestration, and the software/product layer built on top.
Current founder concerns about AI lab roadmaps may reflect focus on immediate exits rather than long-term dedication. Building exceptional products requires decades of commitment, not short timeframes.
Netic seeks remarkable people with agency demonstrated consistently throughout life, not just single examples. The screening process examines whether candidates have started projects, maintained follow-through, and persisted through challenges. The question "What's the hardest thing you've ever done?" reveals how people react to circumstances they can control and change.
A recent hire described maintaining a rigorous health regimen and work commitment for over 15 years without boredom, representing dedication to consistent daily practice rather than grand gestures.
The north star is building an autonomous enterprise where everything except actual service delivery and human labor components is handled autonomously by Netic, enabling companies to focus entirely on delivering customer delight. The company also prioritizes maintaining a culture of only remarkable people, following models like Palantir, SpaceX, and Notion in their long-term craftsmanship approach.
Selling AI solutions and agents to traditional industries like roofing and other service businesses has proven less challenging than expected. The perception that these sectors are old school represents a significant misconception. Some of the most tech-forward business-focused founders have been encountered in these industries, particularly among large enterprises that are extremely value-focused and must be technologically advanced to remain competitive.
Large enterprise contracts can reach half a million dollars and close within 14 days from start to finish. This rapid timeline occurs because companies are extremely thoughtful about their technology investments and rigorously verify that value exists before committing. The buying behavior in these industries demonstrates careful evaluation rather than impulsive adoption.
Traditional industries operate as both extremely primal and technologically sophisticated simultaneously. A large roofing company requires door knockers who go neighborhood to neighborhood assessing roofs and discussing new roofing solutions or solar installations. However, these same companies now utilize comprehensive platforms that handle inbound and outbound communications, analytics, and labor delivery coordination.
Advanced platforms connect to satellite data to analyze how hurricanes affect roofs in different neighborhoods and determine appropriate materials for various conditions. This information feeds autonomously into agent contexts, enabling better customer conversations and more strategic targeting of potential customers. This represents a significant advancement from previous manual processes where companies had to source data externally and analyze it through human review.
Before implementing unified platforms, companies were already attempting to leverage additional data sources but had to sift through information using different tools or manual processes before deploying door knockers. The consolidated approach allows businesses to assist with inbound advertising leads while simultaneously building neighborhood context and reaching out for new revenue generation within a single integrated system.
When private equity firms own 20 to 30 businesses across different industries, unified platforms enable cross-selling opportunities based on customer context. For example, if a customer has a dog and shows interest in wellness offerings, the system can recommend pet care services from portfolio companies, creating more relevant and personalized business development approaches.
The private equity investment playbook has fundamentally shifted from acquiring undervalued companies with high multiples, replacing management teams, and reselling for profit. Since undiscovered high-value opportunities have largely disappeared, the new approach focuses on creating tangible value within existing portfolio companies through technology implementation.
Private equity responses to AI vary significantly. Some firms actively experiment with new technologies and may even deploy companies or partner with specialized AI organizations. Others move more slowly in adoption. The most sophisticated firms now hire dedicated operating partners with AI expertise and maintain teams of AI-forward engineers who educate portfolio companies on implementation strategies.
A critical distinction exists between treating AI platforms like typical software implementations versus understanding that AI relationships require ongoing optimization. Results should improve throughout the year rather than degrade, and companies must separate AI vaporware from solutions that deliver genuine return on investment. The testing timeline should extend beyond initial week-long evaluations to assess sustained performance.
Effective sales processes involve demonstrating live deployments rather than traditional product demos. When platforms are actively generating value for existing customers, showing real-world performance metrics becomes more compelling than theoretical presentations. This approach has proven effective in educating private equity firms about AI capabilities.
One platform has generated over $600 million in revenue for customers through AI-handled interactions. This concrete measurement serves as a powerful demonstration of AI value, showing exactly how real customers interact with the technology and what financial results it produces.
Private equity firms serve more as guidance providers than direct decision-makers for their portfolio companies. The most effective approach involves being a counselor and advisor rather than attempting to impose solutions. This relationship-based strategy acknowledges that private equity cannot simply mandate technology adoption across their holdings.
Initial conversations with private equity firms typically focus on cost reduction because many have not encountered platforms that emphasize revenue generation over expense reduction. The shift toward net new revenue creation represents a significant departure from traditional cost-cutting approaches. AI applications should expand beyond efficiency gains to explore revenue enhancement possibilities.
Personal experiences with limited educational access highlight how technology can democratize learning opportunities. Finding mentors through social media connections with alumni who had moved to different countries provided crucial feedback and guidance. This background informs excitement about AI's potential to provide universal access to education and learning resources.
AI technology removes resource barriers that previously limited access to information and opportunities. The ability to carry comprehensive learning resources in a pocket represents a fundamental shift toward greater individual agency. This democratization of access eliminates previous constraints based on geographic or economic limitations.
While technology makes action easier than ever before, the majority of people may still choose not to take advantage of available opportunities. This creates potential dystopian outcomes where technological advancement benefits only those who actively engage with new possibilities, while others remain behind despite increased accessibility.
Beyond cost-cutting applications, AI can provide meaningful assistance during people's most challenging circumstances. Use cases include helping individuals understand health records, improve sleep patterns, and access support during difficult periods. These applications focus on positive impact rather than job displacement concerns.
Current mainstream AI discussions tend to emphasize negative aspects such as job losses and industry disruption. However, real-world applications demonstrate AI's potential for positive impact, particularly in education and healthcare contexts where the technology provides genuine assistance to individuals in need.
Keep No Priors: AI, Machine Learning, Tech, & Startups in your library
Save the videos and channels worth coming back to, and find them again in one place.
