Addfor is an AI company that has been delivering AI solutions since 2008. It has two divisions - an R&D department that identifies and evaluates advanced AI algorithms, and an application division that industrializes chosen technologies to meet customer needs. Addfor works on problems in various industry segments like energy, automotive, manufacturing, and military, developing solutions like anomaly detection, predictive maintenance, and reinforcement learning applications. It aims to select and deliver the best AI technology for its customers' needs.
2. WE DELIVERTHE BESTTECHNOLOGY
TO OUR CUSTOMERS
Mission
Addfor deliver Artificial Intelligence solutions since 2008
with a single mission:
SELECT and DELIVER the best technology for the needs of our customers.
PRESENTER NOTES
3. WORK PROCESS
We always start from a technical
analysis to check what is
applicable and which could be the
predicted results.
Made on the customer’s specific
request to proof the feasibility of
the solution and to set the right
expectation.
We are focused on mathematically
hard problems: usually our
solutions represent a significant
step forward if compared with the
traditional approaches.
Going in production we help our
customers to choose the right
hardware and the right
environment with scalability in
mind.
We always start from a technical analysis to check what is
applicable and which could be the predicted results.
Made on the customer’s specific request to proof the feasibility
of the solution and to set the right expectation.
We are focused on mathematically hard problems: usually our
solutions represent a significant step forward if compared with
the traditional approaches.
Going in production we help our customers to choose the right
hardware and the right environment with scalability in mind.
PRESENTER NOTES
4. COMBINATORIAL INNOVATION
+
INDUSTRIALISATION
Methodology
We are organised into two divisions:
The first division is the R&D department that is responsible for identifying, evaluating and reviewing all the most advanced Artificial
Intelligence algorithms made available from world’s leading research centres, universities and companies (ex: Facebook, Google …).
Our second division, has an application focus: to industrialise the chosen technologies to meet the specific customers needs. In order
to deliver robust end-to-end integrated solution, we have finalised strategic partnerships with leading hardware manufacturing
companies and suppliers: E4, VI-Grade, NVIDIA and IBM.
PRESENTER NOTES
5. COMBINATORIAL INNOVATION
add-for.com
These are some of the libraries we integrate in our software solutions:
TensorFlow designed by Google
Torch 7 by Facebook
Caffe is more specific for image processing and is optimised by NVIDIA
PRESENTER NOTES
6. INDUSTRIALISATION
TRAINING INFERENCE
add-for.com
We support production training systems like the NVIDIA
DGX-1 and OpenPower.
For inference we support server targets and specific
hardware like the NVIDIA Drive PX2 for Automotive
application, or, alternatively any other kind of device like
mobile platforms.
PRESENTER NOTES
22. GENERAL ANOMALY DETECTOR -TAXI CALLS
In this case we analyse an Open Source DB: the
NYC Taxi Calls: it contains the number of taxi calls
hour by hour in the year 2014.
PRESENTER NOTES
23. ALGORITHM FINDS ANOMALIES
BY ANALYSING RAW DATA
The algorithm has been supplied with only the raw data.
No other data (for example - holidays or special events) has been provided.
Here we see the algorithms that start to analyse the data and after a while are
able to deduct normal daily behaviour and during the peak hours.
When the behaviour differs from the algorithm prediction the point is defined as
an anomaly and is plotted in yellow or red depending of the anomaly strength.
PRESENTER NOTES
24. FIRST ANOMALY: XMAS + NEWYEAR
The first anomalies detected have an easy explanation: on new year’s day
there is a peak of taxi calls just after midnight and on December 25th there
are very few calls.
Remember that the algorithms do not have any information about holidays,
it sees Christmas just as a standard day, for this reason it detects an
abnormal pattern in the data.
PRESENTER NOTES
26. THIRD ANOMALY:
The third anomaly was regarding a low volume of calls the night after
Thanksgiving. It was probably due to a bomb threat spread by the news.
PRESENTER NOTES
27. LAST ANOMALY:WHAT’S APP ON DEC 6 2014 ?
What’s strange is this anomaly detected on December the 6th:
it’s difficult to spot by eye but here the system detected a
strong glitch in data.
Seemingly nothing special happened in NYC that day.
PRESENTER NOTES
28. After having looked back in the NYC News we found this event
that seemingly brought many unexpected visitors to NYC
Summarising, Artificial Intelligence can watch data streams
24/7 finding irregular behaviour, faults, and useful information
to improve your systems and your businesses.
PRESENTER NOTES
29. add-for.com
SPECIALISED ANOMALY DETECTOR - ENERGY
EASUREMENT ANOMALY
STAND-BY POWER
ILLUMINATION
ANOMALY - VENTILATION LEFT ON AT NIGHT
SHOP OPENING HOURS (white box) Our algorithms can detect unusual
conditions in energy meter streams
PRESENTER NOTES