AI Systems Engineer for Time Series Forecasting, Optimization, and LLM-based Coordination (m/f/d) - Early Stage Startup
AI Systems Engineer for Time Series Forecasting, Optimization, and LLM-based Coordination (m/f/d) - Early Stage Startup
AI Systems Engineer for Time Series Forecasting, Optimization, and LLM-based Coordination (m/f/d) - Early Stage Startup
AI Systems Engineer for Time Series Forecasting, Optimization, and LLM-based Coordination (m/f/d) - Early Stage Startup
adjusted flow
Maschinenbau, Betriebstechnik
Karlsruhe
- Art der Beschäftigung: Vollzeit
- 50.000 € – 60.000 € (Unternehmensangabe)
- Vor Ort
- Zu den Ersten gehören
AI Systems Engineer for Time Series Forecasting, Optimization, and LLM-based Coordination (m/f/d) - Early Stage Startup
Über diesen Job
Looking for an opportunity to shape our future sustainably?
Then, adjusted flow is your place to be!
We are developing a software platform for energy consultants and energy-intensive manufacturers to optimize the use of on-site flexibility in terms of generation and consumption..
Our goal: to drive the energy transition where it has the greatest impact—in industry.
We are just getting started and are currently laying the groundwork for a scalable product. To this end, we are looking for an AI and Forecasting Developer (m/f/d) to support us in building and further developing our platform.
Aufgaben
You will work on the AI and decision-logic core of EOIS for industrial energy optimization. The role combines probabilistic forecasting, uncertainty modeling, optimization coupling, and bounded supervisory intelligence for live industrial systems.
-
Develop and improve time-series forecasting models for industrial electricity consumption, PV generation, and other energy-relevant signals
-
Extend deterministic forecast services toward probabilistic outputs such as quantiles, confidence scores, and calibrated uncertainty intervals
-
Evaluate forecast quality on noisy, incomplete, and non-stationary industrial data
-
Design how forecast outputs and uncertainty measures are used in downstream optimization workflows, especially for flexibility assets such as battery storage
-
Contribute to the supervisory intelligence layer that interprets system state, forecast quality, and service health
-
Define and refine controlled decision logic such as re-forecasting, switching operating modes, suppressing low-confidence outputs, or escalating to a human
-
Work on structured AI workflows and production integration into Python services, gRPC interfaces, and the live EOIS platform
-
Support validation, benchmarking, and the technical interpretation of results in real customer environments
Qualifikation
Your Personality Profile
We’re not looking for a perfect checklist—we’re looking for people who think for themselves, ask questions, and are eager to really build something.
- You don’t wait for someone to tell you what to do; you contribute your own ideas
- You question things that aren’t working well and bring your own ideas to the table
- You have no problem disagreeing with us if you’re convinced there are better solutions
- You’re eager to take on responsibility—even if everything isn’t perfectly defined yet...
In terms of technical skills, you should have experience in several of these areas:
Should have skills
- Python
- time-series forecasting for industrial or energy systems
- probabilistic forecasting methods such as conformal prediction, quantile regression, ensemble approaches, or probabilistic boosting
- calibration and uncertainty evaluation
- LP/MILP or related optimization understanding
- Pyomo, PuLP, or equivalent optimization tooling familiarity
- experience designing model outputs for downstream decision systems
- LLM application engineering with structured outputs
- ability to define bounded decision/action spaces
- industrial energy domain understanding, especially PV, BESS, load, and ideally hydro
Helpful skills
- pvlib or equivalent PV modeling tools
- experience with open-source LLM serving ecosystems such as vLLM, Ollama, llama.cpp server
- experience with replay testing / simulation frameworks
- experience with explainability, auditability, or human-in-the-loop AI systems
You don’t have to be an expert in all of these areas. What matters most to us is that you’re willing to dive deep into topics and truly want to understand and implement them.
Benefits
We don’t have an endless list of corporate perks—but we do have an environment where you can really make a difference.
Of course, we still have coffee, snacks, and a great team. We’re open to suggestions for everything else. :)
- A lot of responsibility right from the start
- A steep learning curve—you’ll be doing things you’ve never done before
- Direct influence on the product and decision-making
- Work on a problem that really matters: the energy transition in industry
- Close teamwork and quick decision-making processes
- The opportunity to take on a key role within the company over the long term
We’re currently building the foundation of our company. It’s not always perfectly structured—but that’s exactly why you can make a huge difference here.
If you’re interested in helping us build something, taking on responsibility, and making things better, get in touch.
Candidates should ideally have:
- at least 2 years of total working experience
- including at least 1 year in a small company, startup, or industry research lab
- practical experience in non-state-funded, execution-oriented environments
The key point is that we are not primarily looking for candidates who have only worked in:
- universities,
- publicly funded academic research,
- or very large corporations with highly structured organizations
Those backgrounds are not automatically a problem, but on their own they are usually not enough for our context.
A resume is great—but it’s even better if you send us a quick note explaining why you’re excited about this opportunity and what projects you've worked on where you have used the technical skills we would want you to posess.
The ideal start date is June 1!