Website:
crafter.co.in
Job details:
Role:
We’re building something most tools have ignored for decades:
Not faster communication.
Smarter communication.
Nova AIRA is a communication intelligence system for modern workplaces. It understands how teams actually interact, detects where things slow down or break, and helps teams move with clarity.
This is not a chatbot.
This is not a dashboard product.
We are building a real-time system that understands conversations and turns them into actionable signals.
What You’ll Work On
You will design and build the intelligence layer behind AIRA.
This includes:
• Understanding conversations in real time
Detect intent, clarity, urgency, ownership, and decision signals across threads
• Detecting friction before it becomes a problem
Identify stuck conversations, slow movement, looping discussions, and unclear ownership
• Building intervention systems
Trigger intelligent nudges like follow-ups, ownership clarification, and channel shifts
• Creating manager-level intelligence
Convert raw communication signals into clear intervention points like
“this is stuck”
“this is slowing down”
“this needs attention”
• Designing learning systems
Continuously improve based on behavioral patterns, feedback loops, and usage data
This is not one model.
This is a system of models, logic, and real-time decision layers.
What This Role Requires
We are not looking for a generic ML engineer.
We are looking for someone who can build applied intelligence systems that work in real products.
You should be strong in:
• NLP and LLM systems
Experience with text understanding beyond basic sentiment
Ability to work with context, intent, and multi-message conversations
• Applied ML system design
Comfortable combining LLMs, rules, heuristics, and scoring systems
Ability to design systems, not just train models
• Real-time and event-driven thinking
Experience working with live systems, streaming data, or trigger-based pipelines
• Ranking and scoring systems
Understanding of prioritization, weighting, and decision logic
• Data handling and system thinking
Ability to structure and work with conversational and behavioral data
• Product thinking
You can translate messy real-world behavior into simple, useful outputs
Tech Expectations
You should be comfortable working with:
• Python
• LLM APIs or open-source LLMs
• NLP tooling and pipelines
• Vector databases or embeddings
• Backend APIs (FastAPI or similar)
• SQL and data processing
Stack is flexible. Thinking is not.
Experience
• 5 to 7 years is ideal
• 7 to 10 years is a strong plus
You must have built or worked on real-world ML systems, not just academic projects.
What Matters Most
You understand that:
This is not about predicting text.
This is about understanding human communication.
If you think only in terms of accuracy, precision, and recall, this role will feel limiting.
If you think in terms of behavior, patterns, and decisions, this will feel exciting.
Why This Role Is Different
Most tools make communication faster.
We are making it clearer.
Most systems analyze data after things break.
We detect and fix issues while they are happening.
If done right, this becomes a new layer in workplace software.
How to Reach Out
Send a short note explaining:
• What kind of systems you’ve built
• Why this problem interests you
• One example of a messy real-world problem you’ve simplified
We care more about thinking than buzzwords.
write an email to navatej@novacommunicate.com
Click on Apply to know more.