Junior AI Engineer

At OptiComm.AI, we’re rebuilding global commerce around real customer needs. Our platform predicts what each customer will buy, when, and in what quantity - and turns those predictions into decisions and actions through automation. We operate on a simple loop: Predict → Decide → Act. Predict: forecasting & next-order prediction Decide: prioritize opportunities, risks, and next best actions Act: execute via workflows and AI agents (voice/chat/WhatsApp/email + internal tools) By connecting internal business data with external signals, OptiComm.AI helps retailers, distributors, and manufacturers anticipate demand before it happens, recover lost orders, increase retention, and cut waste. We already have working live customers, a strong pipeline and an ambitious roadmap. Now we're adding a Junior AI Engineer to help grow both our predictive engine and our agentic execution platform. THE ROLE: As a Junior AI Engineer, you'll work on both sides of our platform: the predictive models that forecast orders, detect churn risk, and surface the next best action for every account the AI agents that turn those predictions into conversations and outcomes with real customers This is a role for someone who takes ownership from day one. You'll get clear problems to solve and the freedom to solve them your way. We expect you to dig in, research, try things, and come back with a working solution or a well-reasoned proposal. The team is there when you need it. You're the one driving your work forward. WHAT YOU'LL OWN: Predictive models Build and improve models that predict when customers will reorder, which products they'll buy, and how much Work on churn detection, cross-sell, and next-best-action models that rank accounts by revenue potential and urgency Work with per-customer models that are recalibrated daily, and help improve their accuracy, confidence scoring, and drift detection Explore and engineer features from order history, buying cadence, and basket composition, plus external signals such as seasonality, weather, holidays, local events, and competitor stockouts Run experiments, analyze errors, and draw your own conclusions about what improves performance AI agents Build and improve LLM-based agents that act on predictions through voice, WhatsApp, and email: restock reminders, follow-ups, pre-filled orders, and sales conversations Connect agents to tools and data such as our prediction models, CRM, ERP, and order systems, so they have the right context for every customer Work on the core building blocks of agents: instructions and prompts, tool calling, memory and customer context, and coordination between customer-level and sales-rep-level agents Build evaluation sets and test scenarios, review conversations and outcomes, find where agents fail, and improve them Help implement guardrails that keep agents reliable, transparent, and compliant with GDPR and the EU AI Act Explore ways to improve agents beyond prompting, including fine-tuning and learning from real interaction outcomes Closing the loop Help build the feedback loop where calls, orders, and outcomes flow back into model retraining and agent improvement Contribute to pipelines, monitoring, and workflows that keep models and agents running well in production Work with engineering, product, and client-facing teams to understand what our customers need and turn it into working features WHAT WE'RE LOOKING FOR: Good Python skills and hands-on experience with some of PyTorch, scikit-learn, pandas, and NumPy A solid grasp of ML fundamentals: validation, overfitting, feature engineering, evaluation metrics, and error analysis Practical knowledge of time series, forecasting, or predictive modeling on tabular data, from work, research, or personal projects Hands-on experience building with LLMs, at work or in personal projects. For example: LLM APIs, tool or function calling, RAG, or an agent framework such as LangGraph, CrewAI, or the OpenAI Agents SDK A self-taught mindset. You learn new tools and concepts on your own, from documentation, papers, and experimentation You enjoy solving hard problems and stick with them until they're solved The ability to work independently, make reasonable decisions on your own, and move forward without constant confirmation Clear communication about what you tried, what you found, and what you recommend WHAT WE OFFER: Hands-on work across both predictive ML and AI agents, on real projects for large companies Room to grow your career together with a fast-growing company A friendly team that cares about people and invests in their growth Equity, so you can share in the value you help create NICE TO HAVE: Practical knowledge of voice AI, such as speech-to-text, text-to-speech, or real-time conversational agents Familiarity with LLM evaluation, agent testing, or observability tools Some exposure to fine-tuning LLMs or to reinforcement learning Understanding of Model Context Protocol (MCP) or similar ways of connecting agents to tools Working knowledge of SQL, APIs, or data pipelines, ideally with CRM or ERP data First steps with AWS or another cloud environment Awareness of GDPR and responsible AI practices Interest in sales, B2B distribution, retail, or supply chain JOIN US! If you take ownership from start to finish and care about the outcome, not just the task, we'd love to have you on board. Let's help businesses around the world predict and shape their future, together 🚀