Pronomy

— About Pronomy.io — 

We're building an AI agent that autonomously handles customer support tickets.

Here's how it works in simple terms: companies have knowledge bases (docs, FAQs, past conversations). Our AI learns from this data, then responds to customer questions on its own—no human needed for most tickets.

While it sounds straightforward, the backend involves sophisticated AI architecture that handles complex queries, takes real actions (like rescheduling meetings or updating CRMs), and knows exactly when to escalate to humans.

We've completed multiple startup accelerators (including Ideas Lab and MissionPossible), secured multiple grants, have customers, and have more coming.

— Why This Role Exists — 

Right now, we're talking to customers constantly - in Telegram, email, Slack, on calls. Every conversation has useful insights, but without a system to capture and organize everything, valuable information gets lost.

You'll build the systems that keep our customer operations running - organize customer info, create presentations, prepare documentation, and help onboard new customers. You'll work directly with the founders, see how we make decisions, and get a real look at how early-stage startups operate day-to-day.

You won't be doing this alone. The founder will mentor you throughout - sharing frameworks, recommending resources, and helping you develop the skills you need to succeed.

— What You'll Actually Do — 

1. Customer Operations

• Document every customer conversation (whether it happens in Telegram, email, or calls) in our structured Notion system;
• Research customer companies: their industry, pain points, tech stack, support volume;
• Create company profiles with all relevant information organized in one place;
• Track customer feedback, feature requests, and pain points;
• Onboard new customers: guide them from first conversation to successful product implementation.

2. Documentation & Presentations

• Create customer-facing documentation that explains complex features simply;
• Build presentations tailored to each customer's specific needs;
• Keep all documentation up-to-date as our product evolves.

3. Market Research

• Analyze competitor products and positioning;
• Track industry trends and perform market research;
• Create "what we've accomplished" reports showcasing our progress;

— Requirements — 

1. Must Have:

• Experience creating presentations (university projects, incubator pitches, student organizations - we need to see that you can structure information visually);
• Proficiency with Notion or similar tools for organizing information and documentation;
• Strong analytical skills - you naturally organize information logically, can synthesize insights from multiple sources, and create clear reports;
• Strong written English (B1);

2. Nice to Have:

• Experience creating any projects (even failed attempts count - that's valuable learning);
• Background in research, data analysis, or composing analytical reports;
• Experience creating onboarding documentation or process guides.

— This Role Is Perfect For You If — 

- You want real startup experience—not a resume line, but actual insight into building a company from the ground up;

- You're analytically minded: you enjoy organizing information, researching companies, and finding patterns in data

- You're a fast learner who picks up new tools and concepts quickly, then executes independently

- You're detail-oriented—polished presentations, zero typos, complete information

- You want to learn and grow: The founder will mentor you - sharing startup resources, useful frameworks, and lessons from building companies. You'll get direct guidance on what actually works.

— Time Commitment — 

~20 hours per week with flexible scheduling (you manage your own time);

— Compensation — 

This starts as an unpaid internship. If you deliver strong results in the first 3 months, you'll transition to a paid role as we grow. We're investing time in mentoring you - you're investing time in building real skills and getting insider experience. Both sides win when you prove yourself.

Duration: 3 months initially, with strong potential to convert into a full-time paid position.

— What We're Building — 

At pronomy.io we're building an AI agent that autonomously handles customer support - but our goal isn't just automation. We're building an AI agent that fully mimics and replicates human conversation and behaviour.

Most AI agents today feel robotic. They miss context. They can't detect sarcasm. They respond with templates instead of genuine understanding. We're solving this by embedding scientific foundations - linguistic ontology and cognitive science insights - directly into how our AI thinks and responds.

We've completed multiple startup accelerators, secured multiple grants, have customers with more coming.

We're building this for customers and for the research community. We plan to publish some of our findings and contribute open-source work that advances the field beyond just our product.

— Your First Task — 

Embed Human Communication Mimicry into the AI Agent

This is the fascinating part: teaching an AI model to communicate exactly like humans do (and we aren't just talking about changing prompts, even though that's also an option; it may involve Bayesian state estimation, RLHF, or multi-agent systems).

How do people actually converse? 

When do they ask follow-up questions versus move on? 

How do you detect sarcasm in text? 

What are the natural chat patterns humans follow when they're frustrated, confused, or satisfied? 

Your job is to translate these behavioral insights from cognitive science and linguistics into the model's decision-making framework. Make the AI genuinely understand conversational dynamics, not just pattern-match responses.

— What You'll Actually Do — 

Research & Algorithm Development

• Investigate cognitive science and linguistic approaches to detect ambiguities in unstructured data
• Develop logic and frameworks for knowledge base unification
• Research human communication patterns and translate them into implementable systems
• Design models for conversational context understanding and behavioral mimicry
• Document your research and prepare findings for potential publication or open-source contribution

Implementation & Validation

• Collaborate with the engineering team to embed your research into production systems
• Design experiments to validate your frameworks
• Work on automated data preparation pipelines for customer knowledge bases
• Iterate on AI agent accuracy through your contributions

Note: If you don't have ML or coding experience, that's fine - we'll work together on implementation. Your job is to bring the scientific foundation. If you have ML experience - great, you will have a chance to implement that on your own 😃

— Requirements — 

Must Have:

• Background in NLP, cognitive science, computational linguistics, psycholinguistics, behavioural science, or related field;
• Ability to develop logical frameworks and structured approaches (even if you haven't coded them);
• The capability to translate research concepts into systems that can be implemented

Nice to Have:

• ML experience with RAG systems, LLM fine-tuning, RLHF;
• Experience with dialogue systems or conversational AI;
• Startup or fast-paced research environment experience.

— What Makes This Different —

You'll be joining a small founding team working on foundational AI technology. Your work will directly shape how our AI agent communicates with real customers.

Therefore, you'll have direct influence over the scientific direction, namely Your contribution will be visible in the product, and we're planning to publish research and make some open-source contributions to the field - so your work creates meaningful impact beyond commercial success. We can discuss the specifics of this research contribution in detail.

— Compensation — 

We're resource-constrained as an early-stage startup, so we can't offer competitive big-tech salaries. 

However, we can provide meaningful company options and we're open to discussing monetary compensation that works for both sides. 

More importantly, you'll have real ownership and make genuine impact - both on the product and the field. All the resources (tools, computational resources) of course will be fully covered.