Applied AI Engineer

Role at a Glance

Fully remote Applied AI Engineer at bjakcareer (operating as A1), an Israel-based AI startup. You own AI features end-to-end: model behavior, system design, and production reliability. Core stack: Python, PyTorch/JAX, LLMs, vLLM, vector databases. No salary disclosed. Team prizes fast iteration, independent ownership, and shipping decisions grounded in real-world signals over internal benchmarks.

What This Role Is Really About

Most AI engineering roles hand model outputs to a separate team to productionize. This one doesn't. As an Applied AI Engineer at bjakcareer, you own the entire arc — from shaping model behavior through prompts and agent design, through building the surrounding systems, to sustaining reliable performance in real-world usage. The company is building a proactive AI assistant for everyday digital tasks. You'll operate at the intersection of ML, infrastructure, and product, debugging across abstraction layers and shipping iterations driven by production signals rather than demo-room benchmarks.

About A1 (bjakcareer)

A1 is targeting a straightforward but ambitious observation: the five billion people using email, notes, and task apps daily rely on tools that are fundamentally reactive — they only act when told. A1's product inverts that by building a proactive assistant that handles multi-step tasks, maintains persistent context across sessions, and completes real-world errands with minimal user prompting. Their stated target is roughly a 90% reduction in the time users spend on routine digital work. The team is small and operates with collective decision-making, valuing speed and shipping quality work over process overhead.

What You'll Own

• Build and ship complete AI features spanning model selection, system design, and user-facing behavior
• Design and iterate on prompts, memory systems, tool integrations, and agent workflows
• Transform raw model outputs into structured, predictable, and production-reliable behaviors
• Debug issues across the full stack — model layer, orchestration, infrastructure, and UX
• Optimize AI systems for latency, cost efficiency, and sustained production reliability
• Develop lightweight evaluation frameworks to measure and track real-world model performance
• Collaborate with product and engineering to translate ambiguous requirements into working systems

What You'll Need (Inferred from Listed Tech Stack and Role Scope)

• Proficiency in Python for production AI systems
• Hands-on experience with LLM APIs (OpenAI-style) and open-source models such as LLaMA or Qwen
• Working knowledge of ML inference and serving frameworks (e.g., vLLM or equivalent)
• Experience integrating vector databases in retrieval or persistent-memory contexts
• Ability to own problems end-to-end across model, system, and product layers
• Comfort operating in a small team with high individual autonomy and accountability

Ideal Background (Not All Required)

• Strong grounding in machine learning theory and modern neural network architectures
• Direct experience training, fine-tuning, or deploying ML models in production
• Track record of writing clean, maintainable, production-grade code
• History of crossing abstraction layers — moving fluidly between model, infrastructure, and product concerns
• Effectiveness in ambiguous, fast-moving environments without heavy process support
• Demonstrated bias toward shipping, fast iteration, and improvement driven by real-world data

Skills Breakdown: Must-Have vs. Nice-to-Have

The non-negotiable technical floor is the core stack: Python, PyTorch or JAX, LLM APIs, inference tooling such as vLLM, and vector databases — without these, the day-to-day responsibilities aren't executable. Beyond the stack, A1 prizes cross-layer fluency: the ability to move from a model reasoning failure to an infra bottleneck to a UX issue within the same debugging session. ML training and fine-tuning experience is framed as ideal rather than mandatory, making this role accessible to strong inference-and-deployment engineers as well as those from a training background. The end-to-end ownership and independent execution expected throughout point to mid-to-senior experience as the practical floor.

Salary & Compensation

No compensation range has been disclosed for this role. bjakcareer has not published salary, equity, or benefits data in this listing, and JobForMe does not estimate or infer figures the employer has not provided. If compensation is a deciding factor, raise it early — the company runs a focused 3–4 interview sequence and has stated a commitment to transparency and prompt decisions. For independent benchmarking, Applied AI Engineer roles with comparable LLM and inference scope vary considerably based on seniority, geography, and equity structure.

Location & How You'll Work

This is a fully remote position. bjakcareer is headquartered in Israel, but the listing does not specify whether candidates must be Israel-based or whether international applicants are welcome. No visa sponsorship information has been provided. If you are applying from outside Israel, confirm geographic eligibility and timezone expectations with the hiring team before advancing through the interview stages.